Waste Heat Recovery in Turkish Cement Industry Review of Existing Installations and Assessment of Remaining Potential 2018 Waste Heat Recovery in Turkish Cement Industry Review of Existing Installations and Assessment of Remaining Potential Contents Acronyms .......................................................................................................................................................... 4 Acknowledgements......................................................................................................................................... 5 Executive Summary ......................................................................................................................................... 6 Actual WHR Performance Was Lower than Design Projections, yet Boosted Bottom Lines ...................6 Correct Estimation of Waste Heat Potential, Capacity Utilization, and Cost of Technology Determine Financial Performance ...........................................................................................................................7 Moisture Content of Raw Materials Is a Critical Factor in WHR Performance..............................................7 Remaining WHR Potential Represents a Business Opportunity of $450 million to $630 million...........7 1. Application of Waste Heat Recovery to the Cement Industry ............................................................... 9 The Cement Production Process Is Energy Intensive ........................................................................................9 Waste Heat Recovery for Power Production ........................................................................................................10 Waste Heat Recovery Is a Proven Efficiency Measure for Cement Plants ...................................................11 Steam Rankine Cycle (SRC) Is the Most Common WHR System in Cement ................................................11 Organic Rankine Cycle (ORC) Utilizes Lower Temperature Heat for Power Generation .........................12 Market Interest for WHR Increases as Feasibility of Technology Improves .................................................12 Waste Heat Recovery Feasibility Requires Rigorous Analysis.........................................................................13 Technology Supplier Is Also a Capex Factor........................................................................................................14 Cement Industry in Turkey Is One of the Largest in the World........................................................................14 Cement Production Is the Second Biggest Industrial Energy Consumer in Turkey ...................................15 2. Waste Heat Recovery in the Turkish Cement Industry: Current Status and Project Experience .... 17 Forty Percent of Existing Clinker Capacity Has Implemented WHR ...............................................................17 Turkey’s Cement Industry Has Seen Substantial Adoption of Waste Heat Recovery and Has Further Potential ..........................................................................................................................................18 Survey Coverage is Representative of the Turkish Cement Industry ............................................................18 3. There Is Potential for Additional WHR in the Turkish Cement Industry.............................................. 26 ORC WHR Is Economically Viable in the Turkish Cement Industry ................................................................26 The Turkish Cement Industry Still Has 125 to 230 MW of Feasible WHR Potential ...................................27 4. Key Factors Impacting WHR Feasibility ................................................................................................... 30 WHR Project Economics Are Driven by Factors that Impact the Amount of WHR Power Generated and Affect Project Costs .......................................................................................................................30 WHR Capacity Utilization Is a Key Determinant of Power Generated and Project Economics ...............30 Kiln Operating Hours also Drive WHR Project Performance ............................................................................31 Not surprisingly, WHR Capital Cost (CapEx) Has Significant Impact on Project Economics ....................31 Displaced Electricity Prices Drive Project Savings ..............................................................................................32 Variations in Currency Exchange Rates Affect Project Costs ..........................................................................32 Variations in O&M Costs (Opex) Have Only a Modest Impact on Financial Performance........................33 Reasonable Variations in Auxiliary Power Requirements Have Minimal Impact on Project Performance ..................................................................................................................................................................33 Subsidies and Financing Mechanisms May Improve Feasibility for WHR Investments.............................33 Annex A – Waste Heat Recovery Technologies.......................................................................................... 35 Introduction ....................................................................................................................................................................35 Rankine Cycle ...............................................................................................................................................................35 Steam Rankine Cycle (SRC) .......................................................................................................................................36 Organic Rankine Cycle (ORC) ...................................................................................................................................37 WHR Technology Selection.......................................................................................................................................39 Annex B – Best Practices in WHR Design and Operation ......................................................................... 42 Evaluating Recoverable Waste Heat and Power Generation Potential .........................................................42 Preheater Stages ..........................................................................................................................................................43 Air Cooler Configuration.............................................................................................................................................43 Power Generation Potential ......................................................................................................................................43 Raw Material Moisture Content ................................................................................................................................44 Influence of Dust on WHR ..........................................................................................................................................45 Project Feasibility Analysis ........................................................................................................................................46 WHR Project Risks........................................................................................................................................................46 Design Risk ...............................................................................................................................................................47 Construction Risk ....................................................................................................................................................47 Operational Risk ......................................................................................................................................................48 WHR Maintenance Requirements ............................................................................................................................49 ANNEX C – Key Assumptions for the WHR Financial Model and Performance Analysis ..................... 53 Financial Analysis of Existing WHR Systems ........................................................................................................53 Sensitivity Analysis of ORC WHR for Plants without Existing WHR Systems ...............................................54 Levelized Cost of Electricity ......................................................................................................................................55 Detailed Sensitivity Analysis Results.......................................................................................................................56 ANNEX D – Subsidies and Financing Mechanisms for WHR Investments ............................................. 57 Regional Investment Incentives Scheme ...............................................................................................................57 Incentives ..................................................................................................................................................................57 References .....................................................................................................................................................................58 3 List of Acronyms and Abbreviation BAT Best available technology CSI Cement Sustainability Initiative EBITDA Earnings before interest, taxes, depreciation, and amortization IFC International Finance Corporation IRR Internal rate of return OIZ Organized industrial zone ORC Organic Rankine Cycle O&M Operating and maintenance SRC Steam Rankine Cycle TCMA Turkish Cement Manufacturers Association TL Turkish lira TOE Tons of oil equivalent VAT Value-added tax WHR Waste heat recovery 4 Acknowledgements This report is prepared by IFC team consisting of Alişan Doğan, Barbora Bodnarova, Bruce Hedman, Fatih Avcı, Viera Feckova, Viktoryia Menkova and Yana Gorbatenko. The team would like to thank Alexander Sharabaroff, Alexios Pantelias, Boris Nekrasov, Jeremy Levin, Louis Jude Selvadoray, Luca Giacomo Filippini, Luis Alberto Salomon, Michel Folliet, Sivaram Krishnamoorthy for their valuable contributions, feedback and support throughout the preparation of the report. 5 Executive Summary Turkey is the fourth-largest producer of cement in the world and the largest one in Europe. The industry is the second-biggest consumer of industrial energy in Turkey, with a total consumption of 6426 thousand TOE equivalent in 2016, representing about 6 percent of Turkey’s total energy use. Cement manufacturing is an extremely energy-intensive process, with clinker production alone accounting for more than 90 percent of the total energy used by the industry. However, during the clinker manufacturing process, surplus thermal energy is produced, which can be harvested and converted into power. For example, waste heat from the preheater and clinker-cooler exhausts can be recovered and used to provide low-temperature heating needs in the plant, or used to generate power. Waste heat recovery (WHR) can provide up to 30 percent of a cement plant’s overall electricity needs. Besides, it offers several other benefits, including reduced greenhouse gas emissions and less dependency on external power suppliers. Levelized cost of electricity (LCOE) is a convenient measure of the overall competitiveness of a generation investment compared to other electricity supply options1. Existing WHR projects within Turkey showed that LCOE can even go above grid price caused by higher than projected investment costs or lower than anticipated capacity utilization, underscoring the need for adequate upfront preparation and realistic projections of key project performance and cost parameters in project planning. The Turkish cement industry has been an early adopter of WHR technology. The first WHR installation was commissioned in 2011. By the end of 2016, there were 10 clinker plants operating WHR systems with a total design capacity of 100.7 MW. An additional 34 MW of WHR capacity is in development at four other plants. IFC’s Energy and Water Advisory Team launched a review of WHR investments in the cement sector in Turkey in order to: • Analyze how existing WHR installations performed compared with designed capacity and other installations • Identify lessons learned in implementing and operating WHR systems across the sector • Evaluate potential improvements for enhancing performance • Assess additional investments needed to fully utilize the potential for WHR projects in Turkey’s cement industry 1 Levelized Cost of Electricity (LCOE) represents the per-kWh cost (in discounted real dollars and discounted kWhs) of installing and operating a generating asset over an assumed financial life and duty cycle. Key inputs to calculating LCOE for a project include capital costs, operating and maintenance (O&M) costs, financing costs, and an assumed utilization rate. Cement is the key component of concrete, the world’s most widely used construction material. The industry itself is resource-intensive, involving large amounts of raw materials, energy, labor, and capital. The production process typically involves grinding and blending raw material such as limestone, chalk, shale, clay, and sand with additives such as iron ore. The fine mixture is then fed into a large rotary kiln (cylindrical furnace) where it is heated to about 1,450oC. The high temperature causes the raw materials to react and form a hard granular material called “clinker.” Clinker is cooled and ground with gypsum and other additives to produce cement. 6 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY The team gathered raw data from a representative group of 12 integrated cement plants in Turkey, of which six had installed WHR systems and six had yet to decide on implementing WHR. Data was gathered on the basis of survey questions sent out to cement manufacturing plants. Actual WHR Performance Was Lower than Design Projections, yet Boosted Bottom Lines IFC’s team reviewing the survey responses found that the design capacities of the WHR systems were well within international practice standards of generating between 25 and 45 kWh/ton of clinker. The actual WHR generated was lower than design estimates for five plants, but still within the range of international practice expectations. The performance of a WHR system can fall below its designed performance for a variety of reasons, including seasonal variations in moisture content of raw materials or coal, and the specifics of drying practices adopted at the plant, which reduce available heat to the WHR system. Other factors that contribute to sub-optimal performance include changes in kiln operations that reduce the number of WHR operating hours, extended downtime of the WHR system itself due to operational issues, and oversizing of the WHR system in the initial design. Despite a reduced performance compared with design conditions, the WHR systems surveyed provided, on average, 23 percent of a plant’s total electricity needs, with individual levels ranging from 18 percent to 31 percent. As a result, all six plants realized sizeable savings on their purchased electricity costs, boosting their economic bottom lines. The estimated cost of electricity from the WHR systems, including operating and capital expenses, ranged from $18 to $45 per megawatt-hour, compared with current purchased electricity prices which ranged from $91 to $106 per megawatt-hour2. The estimated savings ranged from $1.4 to $2.7 per ton of clinker. Correct Estimation of Waste Heat Potential, Capacity Utilization, and Cost of Technology Determine Financial Performance Existing project paybacks (or IRRs) for the WHR systems surveyed ranged from 6 to 27 percent, while estimated project paybacks varied between 3.2 to 10.1 years. Key design and operational factors which impact a WHR system’s project economics include sizing, capital cost and capacity utilization. This report illustrates that WHR capacity utilization and kiln operating hours have the greatest impact on a WHR project’s economics because these factors have a direct bearing on the amount of electricity produced by the WHR system. Cost factors such as WHR capital expenses, displaced electricity prices, and variations in currency exchange rates also have a significant impact on project financials. Reasonable variations in WHR operating and maintenance costs and in WHR auxiliary power requirements have a more modest impact on project performance, as resulted form modelling. Moisture Content of Raw Materials Is a Critical Factor in WHR Performance Raw material and fuel moisture content are critical parameters affecting WHR performance, since most plants surveyed use the preheater exhaust to dry raw feed and fuel before they enter the kiln. Using waste heat to dry excessive moisture content in the raw feed or fuel reduces the amount of heat available in the exhaust stream entering the WHR boilers, thus reducing the amount of power produced by the WHR unit. While the average raw material moisture content as reported by the companies is within industry norms, subsequent communication with the survey respondents highlighted that moisture content can vary widely over the course of a year due to variations in weather, changes in raw material sourcing and variations in mining conditions. This can have a significant impact on WHR performance. 2 Estimated cost of electricity for WHR projects including operating expenses only ranged from 9.1 to 14.4 USD/MWh. EXECUTIVE SUMMARY 7 Remaining WHR Potential Represents a Business Opportunity of $450 million to $630 million Out of the remaining WHR investment potential, $340 to $470 million USD of investments are financially feasible under current circumstances, with generation capacities ranging from 125 to 230 MW. Applying international WHR practice standards to the 39 clinker plants that have not implemented WHR results in an estimate of total WHR potential ranging from 158 MW to 283 MW, with a corresponding range of total investment potential of $450 million to $630 million. 8 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY 1. Application of Waste Heat Recovery to the Cement Industry The Cement Production Process Is among its reporting companies was 3,510 MJ and 104 kWh, respectively, although these values can vary greatly Energy Intensive depending on the age and configuration of clinker kilns4. Cement, the binding material that is mixed with an Clinker production is the most energy-intensive stage in aggregate such as sand or gravel and water to form cement production, accounting for more than 90 percent concrete, is the world’s most widely used construction of the total energy use and virtually all of the fuel use in the material. Over three tons of concrete are produced each industry. Clinker is produced by pyro-processing the raw year per person for the entire global population, making materials in large kilns. Three important processes occur it the most widely used manufactured product in the with the raw material mixture during pyro-processing. First, world. Twice as much concrete is used around the globe all moisture is driven off from the materials. Second, the than the total of all other building materials combined, calcium carbonate in limestone dissociates into carbon including wood, steel, plastic, and aluminium – and for dioxide and calcium oxide (free lime); this process is most purposes, none of these other materials can replace called calcination. Third, the lime and other minerals in the concrete in terms of effectiveness, price, or performance. raw materials react to form calcium silicates and calcium The preference for concrete as a building material stems aluminates, which are the main components of clinker. from low manufacturing cost, and its ability to be produced This third step is known as clinkering or sintering. locally from widely available raw materials. Turkey is the fourth largest producer of cement globally, after China, The main kiln type in use throughout the world is the rotary India, and the United States3. kiln (see Figure 1). In rotary kilns, a tube with a diameter of up to eight meters is installed at a three- to four-degree Cement production is a resource-intensive practice angle that rotates one to five times per minute. The kiln involving large amounts of raw materials, energy, labor, is normally fired at the lower end, and the ground raw and capital. Cement is produced from raw materials such material is fed into the top of the kiln, from where it moves as limestone, chalk, shale, clay, and sand. These raw down the tube countercurrent to the flow of gases and materials are quarried, crushed, finely ground, and blended toward the flame end of the kiln. As the raw material passes to the correct chemical composition. Small quantities of through the kiln, it is dried and calcined, then finally enters iron ore, alumina, and other minerals may be added to into the sintering zone. In the sintering (or clinkering) zone, adjust the raw material composition. Typically, the fine raw the combustion gas reaches a temperature of 1,800 to material is fed into a large rotary kiln (cylindrical furnace) 2,000oC. Hot clinker is discharged from the lower end of where it is heated to about 1,450oC. The high temperature the kiln and is immediately cooled in large air coolers to causes the raw materials to react and form a hard nodular ensure clinker quality and to lower the clinker temperature material called “clinker.” Clinker is cooled and ground with to reach handling temperature in downstream equipment. gypsum and other additives to produce cement. Cooled clinker is combined with gypsum and other additives and ground into a fine powder called cement. Cement manufacturing is an extremely energy-intensive Many cement plants include the final cement grinding process – the World Business Council for Sustainable and mixing operation at the site. Others ship some or all Development’s Cement Sustainability Initiative (CSI) of their clinker production to standalone cement-grinding indicates that in 2014 the average thermal energy and plants situated close to markets. electricity consumed to produce one ton of clinker 3 United States Geological Survey, 2017. USGS estimated Turkish 4 GNR Database, 2017, World Business Council for Sustainable cement production in 2016 at 77 million metric tons. The Turkey Cement Development, Cement Sustainability Initiative, http//www.wbcsdcment. Manufacturers Association (TCMA) reported 2016 cement production of org/GNR-2014/. 75.4 million metric tons and clinker production of 67.9 million metric tons, http://www.tcma.org.tr. APPLICATION OF WASTE HEAT RECOVERY TO THE CEMENT INDUSTRY 9 TABLE 1. SPECIFIC THERMAL ENERGY CONSUMPTION more kiln thermal energy compared to the most efficient BY ROTARY KILN TYPE dry kiln (see Table 1). Wet process kilns tend to be older operations. Kiln Type Heat Input (MJ/ton of clinker) Three major variations of dry process systems are used Wet 5,860–6,280 worldwide: long dry kilns without preheaters, suspension Long Dry 4,600 preheater kilns, and preheater/precalciner kilns. In 1-Stage Cyclone Suspension Preheater 4,180 suspension preheater and preheater/precalciner kilns, the early stages of pyro-processing occur in the preheater 2-Stage Cyclone Suspension Preheater 3,770 sections, a series of vertical cyclones (see Figure 1), 4-Stage Cyclone Suspension Preheater 3,550 before materials enter the rotary kiln. As the raw material 4-Stage Cyclone Suspension Preheater plus 3,140 Calciner is passed down through these cyclones, it comes into 5-Stage Cyclone Suspension Preheater plus 3,010 contact with hot exhaust gases moving in the opposite Calciner plus High-Efficiency Cooler direction, and, as a result, heat is transferred from the gas 6-Stage Cyclone Suspension Preheater plus <2,930 to the material. Modern preheater/precalciner kilns also Calciner plus High-Efficiency Cooler are equipped with a precalciner, or a second combustion Source: Based on N. A. Madlool et al., “A Critical Review on Energy Use chamber, positioned between the kiln and preheaters that and Savings in the Cement Industries,” partially calcines the material before it enters the kiln so Renewable and Sustainable Energy Reviews 15, no. 4 (2011): 2,042–60. that the necessary chemical reactions occur more quickly and efficiently. Depending on the drying requirements of Rotary kilns are divided into two groups, dry process the raw material, a kiln may have three to six stages of and wet process, depending on how the raw materials cyclones with increasing heat recovery with each extra are prepared. In wet process systems, raw materials stage. As a result, suspension preheater and preheater/ are fed into the kiln as slurry with a moisture content of precalciner kilns tend to have higher production capacities 30 to 40 percent. Wet process kilns have much higher and greater fuel efficiency compared to other types of fuel requirements due to the amount of water that must systems, as shown in Table 1. be evaporated before calcination can take place. To evaporate the water contained in the slurry, a wet process Waste Heat Recovery for Power kiln requires additional length and nearly 100 percent Production State-of-the-art suspension process rotary kilns include FIGURE 1. ROTARY CEMENT KILN (FIVE-STAGE multi-stage preheaters and pre-calciner to preprocess raw CYCLONE SUSPENSION PREHEATER PLUS CALCINER materials before they enter the kiln, and an air-quench PLUS HIGH-EFFICIENCY COOLER) system to cool the clinker product (clinker cooler). Kiln exhaust streams, from the clinker cooler and the kiln preheater system, contain useful thermal energy that can be converted into power. Typically, the clinker coolers release large amounts of heated air at 250 to 340oC directly into the atmosphere. At the kiln charging side, the 300 to 450oC kiln gas coming off the preheaters is typically used to dry material in the raw mill and/or the coal mill and then sent to electrostatic precipitators or bag filter houses to remove dust before finally being vented into the atmosphere. Although maximizing overall kiln efficiency is paramount, it is also essential to recover remaining waste heat from the preheater exhaust and clinker cooler to provide low-temperature heating or generate electricity. Source: U.S. Department of Energy, “Energy and Emissions Reduction Typically, cement plants do not have significant low- Opportunities for the Cement Industry” (2003). temperature heating requirements, so most waste heat Deployment of CCS in the Cement Industry, Report 2013/19 (2013). 10 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY recovery projects have been for power generation. Waste It should be noted that WHR is an energy efficiency and heat recovery can provide up to 30 percent of a cement cost savings measure that can be retrofitted relatively plant’s overall electricity needs and offers a number of easily on existing kilns lines. Over the long term, the Turkish advantages5,6. cement industry is expected to continue its investment in new technologies and kiln upgrades to remain competitive • Reduces purchased power consumption (or reduces with industry best practices. One example is the use of reliance on captive power plants), which in turn state-of-the-art low pressure cyclone preheater towers reduces operating costs. and precalciners. While the number of preheater stages • Mitigates the impact of future electric price increases for clinker kilns is typically driven by raw material drying • Enhances plant power reliability. requirements, kiln systems with five preheater stages • Improves plant competitive position in the market. and precalciner are considered ‘standard’ technology for • Lowers plant grid energy consumption, reducing today’s modern, dry-process plants. Although the average greenhouse gas emissions (based on credit for thermal energy consumption per ton of clinker for the reduced central station power generation or reduced Turkish cement industry compares favorably with the EU fossil-fired captive power generation at the cement average8, there are still a number of less-efficient kilns plant). operating with three and four preheater stages. Retrofitting existing preheater towers with additional stages of low In dry process cement plants, nearly 40 percent of the pressure drop cyclone preheaters and precalciners can total heat input is available as waste heat from the exit increase production, reduce fuel use and lower emissions, gases of the preheater and clinker cooler. The quantity and is a recommended long term upgrade for most kilns. of heat from preheater exhaust gases ranges from 750 to Preheater tower retrofit costs, however, can be significantly 1,050 MJ per ton of clinker while the quantity of heat from higher than the costs of implementing a WHR project (four the clinker cooler ranges from 330 to 540 MJ per ton of to six times higher), particularly if the existing preheater clinker from the exhaust air of the cooler. The amount of tower foundation must be replaced. Preheater retrofits also waste heat available for recovery depends on a number require a significant amount of downtime for the kiln (8 to of factors including kiln design and production, the 12 weeks), resulting in lost revenues in addition to retrofit number and efficiency of preheater/precalciner stages7, equipment and construction costs9. For these reasons, kiln operation (amount of excess air and air infiltration), preheater upgrades are usually conducted in conjunction configuration of the clinker cooler, the moisture content with major kiln renovations. In the interim, and depending of the raw materials, and the amount of heat required for on age and condition of the kiln, WHR can represent an drying in the raw mill system, solid fuel system for cement economic option to increase energy efficiency and lower mill. A portion of the available waste heat is usually production costs for many cement plants. used to dry raw materials and/or solid fuel, and because raw material drying is important in a cement plant, heat Waste Heat Recovery Is a Proven recovery has limited application for plants with higher raw- Efficiency Measure for Cement Plants material moisture content. Often drying of other materials needed for cement production such as slag or fly ash also Waste heat recovery power systems used for cement kilns requires hot gases from the preheater or cooler; in that operate on the Rankine Cycle10. This thermodynamic cycle case, opportunities for waste heat recovery will be further is the basis for conventional thermal power generating decreased. stations and consists of a heat source (boiler) that converts a liquid working fluid to high-pressure vapor 5 Lawrence Berkeley National Laboratory (LBNL), 2008, “Energy (steam, in a power station) that is then expanded through Efficiency Improvement Opportunities for the Cement Industry,” Worrell, Galitsky, Price, January 2008. a turbogenerator producing power. Low-pressure vapor 6 Environmental Protection Agency (EPA), 2010, “Available and Emerging exhausted from the turbogenerator is condensed back to a Technologies for Reducing Greenhouse Gas Emissions from the Portland 8 Cement Industry,” October 2010. World Business Council for Sustainable Development, Cement 7 The number of preheater stages in a cement plant has significant Sustainability Initiative, Getting Ready Now (GNR) Database, http://www. bearing on the overall thermal energy consumption and waste heat wbcsdcement.org/GNR-2014/index.html. 9 recovery potential. The higher the number of stages, the higher the LBNL, “Guidebook for Using the Tool BEST Cement: Benchmarking and overall thermal energy efficiency of the kiln and the lower the potential Energy Savings Tool for the Cement Industry”, LBNL-1989E, 2008. 10 for waste heat recovery. Selection of the number of preheater stages is The Rankine cycle is a thermodynamic cycle that converts heat into based several factors such as cooler efficiency, restrictions on preheater work. Central station power plants that generate electricity through a tower height, or heat requirements for the mill itself.. high-pressure steam turbine are based on the Rankine cycle.. APPLICATION OF WASTE HEAT RECOVERY TO THE CEMENT INDUSTRY 11 FIGURE 2. TYPICAL WASTE HEAT RECOVERY SYSTEM FIGURE 3. WASTE HEAT RECOVERY SYSTEM USING USING STEAM RANKINE CYCLE (SRC) ORGANIC RANKINE CYCLE (ORC) Source: Adapted from Holcim, 2012–2013 and revised. liquid state, with condensate from the condenser returned the relatively low temperature level of heat from the cooler to the boiler feedwater pump to continue the cycle. Waste (250 to 340oC) limits the efficiency of waste heat recovery heat recovery Rankine cycles can be based on steam or systems in cement kilns to a maximum of 18 to 25 percent13. on an organic compound used as the working fluid11. In each case, the working fluid is vaporized in heat recovery Organic Rankine Cycle (ORC) Utilizes boilers or steam generators (HRSGs) by the hot exhaust Lower Temperature Heat for Power gases from the preheater and hot air from the cooler, and then expanded through a turbine that drives a generator. Generation Organic compounds with better generation efficiencies at Steam Rankine Cycle (SRC) Is the Most lower heat-source temperatures are used as the working fluid in Organic Rankine Cycle (ORC) systems. ORC systems Common WHR System in Cement typically use a high-molecular-mass organic working fluid As shown in Figure 2, in the steam waste heat recovery such as butane or pentane that has a lower boiling point, cycle, the working fluid – water – is first pumped to higher vapor pressure, higher molecular mass, and higher elevated pressure before entering waste heat recovery mass flow compared to water. Together, these features boilers. The water is vaporized into high-pressure steam enable higher turbine efficiencies than those offered by by the hot exhaust from the preheater boiler and clinker a steam system. ORC systems can be used for waste heat cooler boiler and then expanded to lower temperature sources as low as 150oC, whereas steam systems are and pressure in a turbine, generating mechanical power limited to heat sources greater than 260oC. ORC systems that drives an electric generator. The low-pressure steam typically are designed with two heat transfer stages. is then exhausted to a condenser at vacuum conditions, The first stage transfers heat from the waste gases to where the expanded vapor is condensed to low-pressure an intermediate heat transfer fluid (for example, thermal liquid and returned to the feedwater pump and boiler12. transfer oil). The second stage transfers heat from the The steam turbine system is best known from power intermediate heat transfer fluid to the organic working plants. While the electric efficiency of a steam Rankine fluid. A heat exchanger (regenerator or recuperator) may cycle can reach 45 to 46 percent in modern power plants, also be added to the cycle represented in Figure 3, to 11 A third type of Rankine cycle is based on a mixture of water and preheat organic liquid with turbine exhaust vapor before ammonia and is called the Kalina cycle. Application of the Kalina cycle to condenser. ORCs have commonly been used to generate the cement industry is still in the demonstration phase. 12 power in geothermal power plants and, more recently, Institute for Industrial Productivity/International Finance Corporation (IIP/IFC), “Waste Heat Recovery for the Cement Sector: Market and 13 Supplier Analysis” (2014). CSI, “Existing and Potential Technologies for Carbon Emissions Reductions in Indian Cement Industry”. 12 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY in pipeline compressor heat recovery applications in the shown that, in large plants, about 22 to 36 kWh per ton United States. ORC systems have been widely used to of clinker (25 to 30 percent of total power requirements) generate power from biomass systems in Europe. can be generated, depending on kiln configuration and drying requirements. This power is considered sufficient to Market Interest for WHR Increases as operate the kiln section on a sustained basis14. The leading manufacturers of waste heat recovery systems using Feasibility of Technology Improves conventional steam system technology are now marketing The Turkish cement industry was an early adopter of WHR, systems with improved performance due to higher steam with the first installation commissioned in 2011. As of the temperatures and pressures as well as higher component end of 2016, there were 10 clinker plants operating WHR efiiciencies that can reach output levels as high as 45 systems with a total design capacity of 98 MW. Interest in kWh/t of clinker under certain conditions. waste heat recovery among cement industries is mainly driven by: Organic Rankine Cycle systems have started to be applied successfully in the cement industry, especially in recent • Rising prices for power and fuel, particularly where years, in addition to a portfolio of applications such as captive power plants prevail. biomass recovery, geothermal power, and compressor • Concerns about grid power reliability, particularly in stations. developing countries where the electricity supply is often controlled by local, state-owned monopolies Waste Heat Recovery Feasibility and the cost of power can represent up to 25 percent of the cost of cement manufacture. Requires Rigorous Analysis • Industry commitment to and government support The project economics of waste heat power generation and policies for sustainable development. depend on a number of site-specific and project-specific factors, including the following considerations: On the technology supply side, Japanese companies spearheaded the introduction of waste heat recovery • The amount of heat available in waste gases (exhaust power systems in the cement industry and introduced the gas volume and temperature) and conditions of technology to China in 1998. Since then, China has become such gases determine the WHR system’s size, the market leader as technology supplier for steam-cycle potentially its technology (for example, ORCs are waste heat recovery installations, both in the number more applicable to lower-temperature exhaust of systems installed domestically and in the number of streams and lower gas volumes), and its overall systems installed internationally by Chinese companies generation efficiency (including the amount of power (particularly in Asia). Initially, waste heat recovery that can be produced). The amount of heat available development in China was driven by incentives such as and at what temperature is determined by the size tax breaks and Clean Development Mechanism (CDM) and configuration of the kiln (that is, its capacity in revenues for emissions reductions from clean energy tons per day and number of preheater/precalciner projects. In 2011 a national energy efficiency regulation stages) and the raw material moisture level (which mandated waste heat recovery on all new clinker lines determines the percentage of hot exhaust gases built after January 2011. These drivers were reinforced needed for drying). when multiple Chinese waste heat recovery suppliers • Capital cost of the heat recovery system, which entered the market, lowering waste heat recovery capital generally depends on size, technology used, and and installation costs by adopting domestic components equipment supplier. and design capability, which developed the technology • System installation costs (design, engineering, for the Chinese market, and by 2012 over 700 units were construction, commissioning, and training) depend operating in that country. The bulk of recent market on the installation size, technology, complexity, activity has been in Asia, where Chinese companies or supplier, and degree of local content. joint ventures are the primary suppliers. The experience 14 CSI, “Existing and Potential Technologies for Carbon Emissions of China in using waste heat recovery for power has Reductions in Indian Cement Industry”. APPLICATION OF WASTE HEAT RECOVERY TO THE CEMENT INDUSTRY 13 • System operating and maintenance costs are trained labor force and prevailing labor rates. Total capital affected by size, technology, site-specific operational cost (equipment and installation) is strongly influenced by constraints or requirements. They are also influenced size – smaller WHR systems will have a higher cost per by staffing – whether the system can be handled by kW of generation capacity. Engineering, civil work, and existing operating staff, new staff that require training, construction costs can represent as much as 34 to 45 or be outsourced. percent of total project cost. Costs in Western countries • Operating hours of the kiln and availability of the heat are at the high end of the range16. Figure 4 shows industry recovery system. estimates of total installed costs for cement WHR projects • Displaced power prices based on grid electricity on a $/kWe basis and illustrates how costs depend heavily no longer purchased, or reduced dependence on on project size (MW), local cost variations (region of the captive power plants and associated costs. installation), and type of technology (systems lower than 2 • Net power output of the WHR system. Net output is to 3 MW tend to be ORC systems). Hence, total installed more important in determining project economics costs for WHR systems are affected by all of the factors than gross power output. The impact of auxiliary mentioned above, but costs can range from 7,000 $/kWe power consumption and process/booster fans must for 2-MW systems (ORC) in Europe to 2,000 $/kWe for be included in efficiency and economic calculations. 25-MW systems (steam) in Asia. IFC’s experience over the • Availability of space in close proximity to the last few years is that costs can be lower than indicated in preheater, cooler, and air-cooled condensers. Figure 4, as also demonstrated by more recent experience • Availability of water. in the Turkish cement industry (see Figure 16). Note that since 2013, the leading suppliers of waste heat recovery A WHR installation is a relatively complex system with systems using conventional steam circuit technology have multiple interrelated subsystems. The basic package for now been marketing second-generation systems with a steam-based system15 consists of heat recovery boilers improved efficiencies, resulting in somewhat lower costs or heat exchangers, steam turbine, gearbox, electric on a $/kW basis. generator, condenser, steam and condensate piping, 16 Holcim, 2013, “Experience and Challenges in Waste Heat Recovery,” lubrication and cooling systems, water-treatment system, Urs Herzog, Thomas Lamare, 2nd Global CemPower, London, June 2013 electrical interconnection equipment and controls. Total installed costs, which include design, engineering, construction, and commissioning, can vary significantly FIGURE 5 – INSTALLED COSTS FOR CHINESE WHR depending on the scope of plant equipment, country, SYSTEMS (STEAM CYCLE)* geographical area within a country, competitive market conditions, special site requirements, and availability of a 15 The discussion of system costs and project economics focuses primarily on steam systems, which represent the vast majority of installed technology – conventional steam systems account for 99 percent of existing WHR installations in the cement industry worldwide. FIGURE 4. WHR INSTALLED COSTS, USD/KW Source: Holcim 2013, OneStone Research 2013; IFC 2014. * The above capex estimates are based on Chinese WHR equipment. Experience from WHR project in various regions suggest that installation costs are often higher, and in certain cases reach up to $5,000 per kWe depending upon WHR power technology type and installed capacity. For instance, European-manufactured WHR power systems could cost up to Source: Holcim 2013, OneStone Research 2012, 2013. $3,800 per kWe. 14 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY Technology Supplier Is Also a Capex in 1953, when the Turkish Cement Industry Company (ÇISAN), a public enterprise, was set up to commission Factor 15 new cement plants throughout Turkey. A total of 17 In addition to factors discussed above, the supplier is a key more plants were added between 1963 and 1980 by the determinant in total capital cost for steam WHR systems in national and regional governments to support regional the cement sector. Due to its extensive domestic installed development. The Turkish Manufacturer’s Association base, China is by far the major player in WHR for the cement (TCMA) was formed in 1957 to represent the interests of industry in terms of installations, equipment supply, and the growing industry. In 1989 cement industry privatization developer experience. Faced with near market saturation began in the west of Turkey, where greater demand at home, and building on the advances Chinese firms have and higher efficiency meant plants were more likely to made in the global cement market, Chinese WHR suppliers be attractive acquisition targets. Plants in the east were and developers are actively marketing WHR systems in restructured and consolidated prior to privatization, which Asia and branching out into Africa, the Middle East, and occurred rapidly by 1997. By this time, Turkey was the other regions. Initially, Chinese suppliers faced concerns third-largest cement producer in Europe after Germany about the reliability and quality of some of their WHR and Italy. By the end of the privatization process, Turkey systems, as well as their ability to provide adequate start- had 40 cement plants producing a total of 33.3 Mt/yr of up and training support17. Nevertheless, Chinese suppliers cement; eight plants were wet process and the rest were are active in a number of countries, including Turkey, dry kilns. Driven by an expanding economy, the Turkish where they are establishing partnerships and alliances with cement industry more than doubled in size in just 15 years; national resources for marketing and local project scope. 33 percent of all Turkish capacity in 2010 was less than six Key to the Chinese success is their commanding price years old20. advantage over Western suppliers. Figure 5 provides an estimate of total project costs for Chinese WHR systems By the end of 2017, the Turkish cement industry has a installed in China, other parts of Asia, and Europe. Note total of 54 active integrated cement plants with 83.6 Mt/ that while the figure depicts relative cost differences for yr of clinker production capacity and 14 cement grinding Chinese WHR systems across these three regions, the facilities with 8.1 Mt/yr of grinding capacity.19 An additional range of reported total installation costs varies widely and three integrated plants are under construction with 5.5 Mt/ can be impacted by a variety of site- and project-specific yr of clinker capacity, and an existing plant is expanding factors13,18. with an additional 1.0 Mt/yr clinker capacity. Table 2 lists the largest clinker producers in 2017. The majority Main SRC system suppliers other than China are from of cement producers are Turkish companies, including Japan, India, and Denmark. ORC systems, on the other Akçansa Çimento, Oyak Group, Çimsa Çimento, Askale hand, are mainly supplied by US, Japanese, and Swiss Çimento, and Limak Çimento. Multinational producers companies. There are hundreds of successful ORC include Heidelberg (a joint venture in Akçansa), Votorantim references in many types of applications throughout Cimentos, Cementir Holding, Titan and Vicat. the world, and waste heat recovery in cement is rapidly becoming one of them. Domestic cement sales were 72.2 million tons in 201721. Domestic consumption is forecast to increase in the near Cement Industry in Turkey Is One of the term with the undertaking of new infrastructure projects, Largest in the World including highways, stadium construction, and new metro lines. Turkey is also constructing one of the world’s largest The Turkish cement industry is the fourth largest in the airports in Istanbul. As noted earlier, Turkish producers are world, and the largest producer in Europe, with reported expanding existing plants to meet planned consumption 2017 production of 80.5 million tons of cement and 70.8 levels. Turkey is a key exporter of cement, shipping 7.9 million tons of clinker19. The industry began in 1911 with million tons of cement and 4.9 million tons of clinker to 20,000 tpy of capacity and saw large development starting over 80 countries in 2017, including Libya, Iraq, Russia, 17 IFC 2014, “Waste Heat Recovery for the Cement Sector”. 20 International Cement Review (ICR), 2103, “The Global Cement Report” 18 For example, Sinoma Energy Conservation Ltd. estimates the costs of 10th Edition, March 2013. a 9 MW system installed on a 5,000 tpd kiln in Asia (outside China) to be 21 TCMA, “Cement Sector Waste Heat Recovery,” June 2017. $18 million to $19 million, or about $2000/kW (Sinoma 2013). 19 TCMA , http://www.tcma.org.tr. APPLICATION OF WASTE HEAT RECOVERY TO THE CEMENT INDUSTRY 15 and Israel as well as growing markets in Africa and TABLE 2. MAJOR CEMENT COMPANIES – INTEGRATED Europe. Turkey is also one of the world’s most significant PRODUCERS (2017)* producers of white cement, produced by Cimsa Cimento Group # of Clinker plants Capacity in Mersin (1.10 Mt/yr) and Adana Cimento (Oyak Group) in Adana (0.35 Mt/yr)22. OYAK 6 10.05 Akcansa 3 7.03 Cement Production Is the Second Limak 7 6.60 Biggest Industrial Energy Consumer in CIMSA 5 5.70 Turkey Askale 5 5.20 Nuh 1 4.40 The cement industry is one of the largest energy consumers in the Turkish economy, with a total consumption of 6426 AS Cimento 1 4.30 thousand TOE in 2016, representing about 6 percent of Cementir 4 4.22 Turkey’s total energy use23. Energy comprises a significant Votorantim 4 3.88 portion of the cost of producing cement, and the industry Medcem 1 3.50 continues to make efforts to reduce energy use to control Vicat 2 3.20 costs and meet national sustainability and environmental goals. Waste heat recovery can be an important factor in Sanko 2 3.10 reducing energy use, costs, and greenhouse gas emissions Kipas 1 2.90 at cement plants by utilizing waste energy normally vented Bati Anadolu 2 2.40 into the atmosphere to produce up to 25 to 30 percent of Goltas 1 2.20 a cement plant’s electricity needs. Others 9 14.9 22 TOTAL 54 83.6 Egypt is the world’s largest producer of white cement, followed by Spain and Turkey. In 2014 there were 31 global producers of white * IFC. cement operating 45 integrated plants in 29 countries with a combined production capacity of 13.3 Mt/yr. Çimsa Çimento, part of Hacı Ömer Sabancı Holding A.Ş., was the second-largest white cement producer in 2014 with its plant in Mersin, and exports its white cement to more than 60 countries around the world. “White Cement Report,” Global Cement, Nov. 2015. 23 TCMA, “Cement Sector Waste Heat Recovery,” June 2017. 16 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY 2. Waste Heat Recovery in the Turkish Cement Industry: Current Status and Project Experience Forty Percent of Existing Clinker Capacity implemented or are in the process of installing WHR to date account for 37 percent of Turkey’s total clinker Has Implemented WHR production capacity, and 88 percent of clinker capacity at Efficiency and productivity improvements remain very plants 1.0 Mt/yr or greater. important for the Turkish cement sector, and the industry was an active early adopter of WHR. As shown in Table 3, To date, all the WHR units installed or currently in the 100.7 MW of WHR capacity was installed during 2011–2015 investment phase have been steam cycle systems. As at 10 plants, representing 21.7 Mt/yr of clinker capacity such, the waste heat recovery power market for the (26 percent of Turkey’s total clinker capacity of 83.6 Mt/ Turkish cement industry has been served primarily by yr). Another 34 MW of WHR capacity is currently in the Chinese and Chinese/Japanese joint venture suppliers, construction phase at four additional plants, representing which have extensive experience in steam-cycle WHR 8.9 Mt/yr of clinker capacity (an additional 11 percent of applied to the cement industry24. Some of the suppliers total clinker capacity). Together, the 14 plants that have 24 Institute for Industrial Productivity/International Finance Corporation (IIP/IFC), “Waste Heat Recovery for the Cement Sector: Market and Supplier Analysis” (2014). TABLE 3. EXISTING WHR SYSTEMS IN TURKEY Plant Clinker Number Commis- Technology Provider WHR Capacity Capacity of Lines# sioned (Mt/yr) 1 Akcansa Çimento - Canakkale 4.45 2 2011 Sinoma EC 15.2 MW 2 Aşkale Çimento - Erzurum 1.95 2 2011 AVIC / GTE Endüstri 7.5 MW 3 Cimsa Çimento - Mersin 2.47 2 2012 Anhui Conch / Kawasaki / 9.8 MW Marubeni 4 Bati Çimento Anadolu 1.40 2 2012 Sinoma EC 9.0 MW 5 Bati Soke Çimento 1.00 1 2012 Sinoma EC 5.5 MW 6 Nuh Çimento 4.40 3 2013 Sinoma EC 17.7 MW 7 Bursa Çimento - Kestel 1.40 2 2013 STEC/Mitsubishi 9.0 MW 8 Bolu Çimento - Oyak Group 1.20 1 2014 STEC/Mitsubishi 7.5 MW 9 Aslan Çimento - Oyak Group 1.25 1 2014 STEC/Mitsubishi 7.5 MW 10 Goltas Çimento 2.17 2 2015 AVIC / GTE Endüstri 12.0 MW 11 Aşkale Çimento – Van* 1.05 1 line - Feasibility Done – Investment 7.5 MW decision is pending 12 Aşkale Çimento – Gumushane* 1.40 1 line - Feasibility Done – Investment 7.5 MW decision is pending 13 KCS Çimento* 2.90 2 lines - Commissioning in Feb 2018 9.0 MW 14 Medcem Çimento* 3.50 1 line - Commissioning in 2018 10.0 MW Source: Based on spesific company inputs and company annual reports. * Investment Phase. # Number of clinker lines incorporated into WHR system. WASTE HEAT RECOVERY IN THE TURKISH CEMENT INDUSTRY: CURRENT STATUS AND PROJECT EXPERIENCE 17 have partnered with Turkish firms for local engineering The team had three specific objectives for the review: support and installation. Major participants include: • To conduct a post-investment review of existing • Anhui Conch / Kawasaki Engineering is a joint venture WHR installations to evaluate their current state and of the Chinese cement company Anhui Conch and operational performance. the Japanese equipment and engineering company • To identify possible bottlenecks for capacity utilization Kawasaki Plant Systems. Anhui Conch / Kawasaki is and adoption of small- to mid-size installations (less a leading WHR supplier in China and has installed than 5 MW) in the cement sector in Turkey and a number of systems in other countries – including • To evaluate the remaining economically viable India, Pakistan, and Vietnam – and one system in potential for steam-cycle and ORC WHR installations Turkey. in Turkey’s cement sector. • Sinoma Energy Conservation (Sinoma EC) is a leading Chinese supplier of waste heat recovery As part of this review, the IFC Team gathered raw data from power generation systems. Sinoma EC has also a representative group of 12 integrated cement plants in installed over 20 WHR systems in other countries – Turkey – of which six had already installed WHR systems, including Vietnam, The Philippines, India, Pakistan, and six had not yet made a decision on implementing Thailand, Angola, the UAE, and Saudi Arabia – and WHR. The survey sample was selected based on broad four systems in Turkey. Sinoma EC has partnered outreach to plants that had operational WHR systems in with SC Endustri AS for WHR applications in Turkey. 2015 and companies with plants producing 1 MT/yr or more • STEC (Shanghai Triumph Energy Conservation) that were not currently utilizing WHR. The survey sought to / Mitsubishi is a joint venture of China Triumph gather information at the plant level on plant operations International Engineering Company (CTIEC) and relevant to WHR potential, such as design/operating Mitsubishi Corporation. Shanghai Triumph specializes capacity, kiln operation data, waste heat parameters, in medium- and low-temperature flue-gas waste heat moisture content of raw materials/fuels, current energy recovery for power generation from glass and cement use, and alternative fuels. For plants that had already kilns. As of 2013 the company had 28 EPC projects installed WHR, the survey requested information on WHR in production, primarily in China and including three system type, design characteristics, actual performance, systems in Turkey. project costs and financing, project implementation, and • SinoPES International Engineering Company is a overall satisfaction with the project development process Chinese engineering company that focuses on and WHR installation. For plants that had not installed cement plant design, engineering, procurement, WHR, the survey requested information on the level of construction, commissioning, operation and familiarity or experience with WHR. The survey ended maintenance, and waste heat recovery. SinoPES with questions for all participants on priorities for future has partnered with GTE Endüstri Sistemleri for WHR efficiency and productivity investments. applications in Turkey. Survey Coverage is Representative of Turkey’s Cement Industry Has Seen the Turkish Cement Industry Substantial Adoption of Waste Heat Surveyed Plants Represent a Larger than Statistically Recovery and Has Further Potential Relevant Proportion of Industry Capacity IFC’s Energy and Water Advisory Team25 recently launched The plants participating in the WHR survey represent a a review of WHR investments to analyze the operational total clinker production capacity of 21.2 Mt/yr, accounting and financial performance of WHR installations, identify for 25.3 percent of the total 83.6 Mt/yr clinker production lessons learned during implementation/operation, and capacity for the industry. Survey coverage by clinker assess what is needed to utilize the full potential for WHR. capacity ranges is presented in Figure 6 below and reflects 25 the survey’s wide representation of the Turkish cement IFC, a member of the World Bank Group, is the largest global development institution focused on the private sector in developing industry. Design clinker capacities of the 12 surveyed countries. IFC’s Energy and Water Advisory Team works with private and plants ranged from 0.435 Mt/yr to over 4.4 MT/yr and, as sub-national clients to facilitate adaption of clean energy and resource efficiency solutions. shown in the figure, covered roughly 20 percent or more 18 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY of each capacity segment of the industry. Evaluating a FIGURE 6. WHR SURVEY COVERAGE VS CLINKER variety of plant-level performance parameters indicates CAPACITY that the 12 surveyed plants are statistically relevant and representative of the Turkish cement industry: Clinker Utilization of Surveyed Plants Is in Line with Overall Sector As shown in Figure 7, reported annual capacity utilization of the surveyed plants for 2015 ranged from a low of 87 percent to a high of 130 percent based on actual production versus maximum design capacity, and from 87 percent to 99 percent based on planned operating hours. The production-based utilization rate for the Turkish cement industry as a whole was 85 percent in 2015 and 88.7 percent in 201626. Variation in Moisture Content Affects WHR FIGURE 7. REPORTED ANNUAL CLINKER CAPACITY Performance UTILIZATION, 2015 Raw material and fuel moisture content are critical factors affecting WHR performance, since most plants use the preheater exhaust to dry raw feed and fuel before they enter the kiln. Using waste heat to dry excessive moisture content in the raw feed or fuel reduces the amount of heat available in the exhaust stream entering the WHR boilers, reducing the amount of power produced by the WHR unit. All 12 of the surveyed plants used preheater exhaust, and sometimes clinker cooler exhaust, to dry raw material. Figure 8 shows the reported average raw material moisture content on an annual basis for 11 of the survey plants, ranging from a low of 3 percent to a high of 10 percent. While the average raw material moisture content as reported by the companies is within industry norms, subsequent communication with the survey respondents FIGURE 8. REPORTED RAW MATERIAL AVERAGE highlighted that moisture content can vary widely over the MOISTURE CONTENT course of a year due to variations in weather, changes in raw material sourcing, and variations in mining conditions. According to discussions with cement plant managers, seasonal or periodic variations in moisture content can have a significant impact on WHR performance. Thermal Energy Use of the Plants Is Representative Figure 9 shows the average thermal energy consumption per ton of clinker produced for each of the survey plants. It also includes the Best Available Technology (BAT) average, which is approximately 3.0 GJ/ton, and 26 TCMA. WASTE HEAT RECOVERY IN THE TURKISH CEMENT INDUSTRY: CURRENT STATUS AND PROJECT EXPERIENCE 19 the 2011 average for the entire Turkish cement industry, FIGURE 9. REPORTED THERMAL ENERGY REQUIRED which was just below 3.5 GJ/ton. In addition, the World PER TON CLINKER PRODUCED Business Council for Sustainable Development’s Cement Sustainability Initiative reports the 2014 EU average thermal energy consumption as 3.51 GJ/ton27. The reported thermal energy consumption of all 12 of the survey plants were higher than the BAT average, but in line with the 2011 Turkish and 2014 CSI EU averages. Seven of the plants had kilns with four preheater stages, and five plants had at least one kiln with five preheater stages. All 12 plants used coal as the primary fuel for their kilns, and seven of them supplemented coal with natural gas firing. In comparing Figures 6 and 7, note that there does not appear to be a direct correlation between thermal energy use and reported raw material moisture content in the survey plants. Raw material is typically dried to FIGURE 10. ELECTRICITY CONSUMPTION FOR CEMENT (KWH/TON CEMENT) specifications before entering the kiln, either through direct fuel dryers or the use of waste heat, so moisture content in the as-delivered raw material is not a factor in kiln operation. As noted above, all survey plants use a portion of the waste heat for drying. While the use of waste heat for drying has no direct impact on kiln thermal energy use itself, it does impact the overall plant fuel use – waste heat used for drying eliminates or reduces the need for separate fuel-fired dryers and would affect overall plant thermal performance accordingly. As noted earlier, the use of waste heat for drying also impacts the performance of WHR systems, reducing the amount of heat available for power generation. Electricity Requirements Vary Based on Individual Plant Operations Correct Assessment of Design Capacity Is Key for Project Financial Performance Figure 10 shows the total electricity requirements of each plant, ranging from 85 to 140 kWh/ton of cement. Electricity An initial modeling exercise estimated the financial requirements in terms of kWh per ton of cement can vary performance of the existing WHR systems included in the widely based on plant specific parameters such as the industry survey. The clinker capacities of the six plants with amount of raw material and finishing grinding conducted on- WHR ranged from 0.61 to 4.45 Mt/yr. The WHR systems site, the grinding technologies used, and the type of cement were all steam Rankine cycles (SRC) installed between itself. The CSI 2014 EU average electricity use is reported as 2011 and 2014, and ranged in design capacity from 5.5 to 118 kWh/ton of cement, with an overall range of 40 to 180 17.7 MW. Two financial analysis cases were evaluated for kWh/ton. Figure 10 also shows that for those plants that had these existing projects: already implemented WHR, the WHR systems on average provided 23 percent of total plant electricity needs, with 1. Planned Performance – Design conditions based on individual levels ranging from 18 to 31 percent. the maximum planned operating hours of clinker lines in a given year with 100% WHR capacity utilization. 27 World Business Council for Sustainable Development, Cement Sustainability Initiative, Getting Ready Now (GNR) Database, http://www. wbcsdcement.org/GNR-2014/index.html. 20 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY 2. Actual Performance – Operations based on reported FIGURE 11. WHR DESIGN AND WHR ACTUAL actual operating hours of clinker lines and actual PRODUCTION VS CLINKER DESIGN CAPACITY WHR capacity utilization in a given year. The financial performance analysis was based on data reported by the individual companies for their operations in 2015. No adjustments were made to primary data received. In case no data was provided for select questions, estimates were made based on other relevant plant parameters and responses of comparable companies within the survey sample. Key assumptions used in the financial analysis are included in Annex C. The results of the analysis summarized below stress that correct assessment of available waste heat at the design stage, followed by operation at high-capacity utilization, are key parameters for better return on investment. Actual WHR Performance Was Lower than Design Performance, but Still Boosted Bottom Lines FIGURE 12. PLANNED AND ACTUAL LEVELIZED COST OF ELECTRICITY (LCOE) As shown in Figure 11, the WHR design capacities of the existing WHR systems were all within international practice standards of generating between 25 and 45 kWh/ton of clinker. Actual WHR generation, however, was lower than design estimates for five of the plants, though still within the range of international practice expectations. Design conditions are based on upfront projections of available waste heat, annual kiln and WHR operating hours, current and future electricity prices, and project capital costs. Actual WHR system performance can fall below design performance for a variety of reasons, including variations in moisture content of raw materials or coal, the specifics of drying practice at a given plant that may reduce available heat to the WHR system, changes in kiln operations that reduce the number of WHR operating hours, extended page 79) represents the per-kWh cost (in discounted real downtime on the WHR sys tem itself due to operational dollars and discounted kWhs) of installing and operating a issues, or oversizing of the WHR system in the initial design. generating asset over an assumed financial life and duty Actual project economics can vary from design conditions cycle or capacity utilization. For technologies such as solar due to unexpected changes in cost factors, including or WHR generation that have no fuel costs and relatively displaced electricity costs or project capital expenses. low O&M costs, LCOE changes in rough proportion to Potential performance and cost uncertainties underscore the capital cost and utilization rate of the system. As with the need for adequate upfront preparation and project any projection, there is uncertainty about these factors analysis based on longer-term measurements and realistic and their values can vary with unanticipated changes in projections of key performance and cost parameters. project operating parameters or project costs. Figure 12 shows the planned and actual LCOE’s of the existing WHR Levelized cost of electricity (LCOE) is often used as a systems. The figure shows the influence of capital costs convenient measure of the overall competitiveness of on LCOE, and highlights the differences between planned a generation investment compared to other electricity and actual LCOEs for the six systems, primarily impacted supply options. LCOE (as described in detail in Annex C, by actual capacity utilization compared to planned capacity WASTE HEAT RECOVERY IN THE TURKISH CEMENT INDUSTRY: CURRENT STATUS AND PROJECT EXPERIENCE 21 utilization for each of the projects. Planned LCOE’s were capital expenses, ranged from 18 to 45 $/MWh, compared all well below the average delivered grid price of electricity to current purchased electricity prices which range from 91 of 96 USD/MWh. Actual LCOE’s were consistently higher to 106 $/MWh28. than planned, with four of the six LCOE’s still well below the grid price of electricity, again underscoring the need Actual Project Paybacks Were Longer than Planned for adequate upfront preparation and realistic projections Paybacks due to Reduced WHR Capacity Utilization of key performance and cost parameters. and Lower-than-Expected Grid Electricity Costs Figure 14 provides a summary of the projects’ financial Even with the reduced performance of actual operation performance based on estimated simple payback29 of compared to design conditions, the WHR systems on the six existing WHR systems for both the planned and average provided 23 percent of total plant electricity actual cases. As shown in the figure, planned system needs, with individual levels ranging from 18 percent to paybacks ranged from 2.2 years to just over six years, 31 percent (Figure 13). As a result, all six plants realized driven primarily by differences in system capital costs. sizeable savings on their purchased electricity costs, The actual case paybacks are based on the reported boosting their economic bottom lines – the estimated cost gross electricity generated at each plant, and estimates of electricity for the WHR systems, including operating and range from 3.1 to 10.1 years – 40 to 70 percent longer than planned paybacks under design conditions. The FIGURE 13. WHR GENERATION AS A SHARE OF TOTAL difference in financial performance between planned and PLANT ELECTRICITY USE actual is primarily due to the reduced capacity utilization of the WHR units (as shown in Figure 13) and grid electricity prices that were lower than originally projected. Estimated payback for one plant that reported electricity prices 30 percent below design projections was 50 percent longer than the planned design payback. High WHR Capacity Utilization Is Key to Project Success WHR system capacity utilization is a key performance parameter for project returns, and can be affected by a combination of factors including kiln utilization, kiln operation and raw material and fuel moisture content. Figure 15 presents the actual capacity utilization factors for WHR systems based on reported gross generation FIGURE 14. SIMPLE PAYBACK – ESTIMATED ACTUAL VS ESTIMATED PLANNED values for each plant. As shown, actual WHR utilization ranged from a high of 94 percent to a low of 63 percent with an average of 73 percent, even though clinker production capacity utilization for all plants was relatively high (87 percent to 130 percent of planned kiln production capacity). WHR system availability for all installations was reported to be in the 95+ percent range, indicating that the lower WHR utilization rates shown in Figure 15 were not due to the systems being unable to operate when 28 The estimated cost of electricity for WHR projects includes operating costs and the 20-year net present value of capex expenses. The estimated cost of electricity including operating expenses only ranged from 9.1 to 14.4 $/MWh. 29 Paybacks presented are simple paybacks for the WHR system using a five-year average (2009–2012) of displaced electricity costs. 22 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY needed30. Low utilization rates are primarily as a result of of economies of scale (larger systems applied to larger higher than expected raw material/fuel moisture content, kilns generally have lower per-kW costs), site-specific especially in winter due to seasonal variations, which installation issues and differences in system suppliers can leads to more heat used for drying and less heat available significantly impact total investment requirements. for power generation. However, as shown in Figure 15, there is little correlation of the WHR utilization rates with The existing WHR systems included in the survey were either kiln capacity utilization or average raw material all commissioned between 2011 and 2014 and reflect moisture content over the year. This leaves the possibility prevailing installed costs of systems in that time period. that the WHR systems may have been oversized in the Anecdotal information from more recent projects indicate original designs, or there may be site-specific conditions that WHR investment costs are decreasing in Turkey. such as seasonal variations in raw material moisture Figure 17 shows reported investment costs for the six content that reduced the amount of heat available for survey installations and two estimates from recent WHR power generation for certain periods. Other site-specific installations currently in the construction phase. Reported operational issues may also impact WHR performance.. As investment costs for the two recent installations are 20 to an example, Turkey is a leading supplier of white cement 25 percent lower than the average cost line for surveyed on the world market, and one of the survey plants is a key provider of this product. White cement production has different thermal and kiln residence time requirements FIGURE 15. WHR CAPACITY UTILIZATION than gray cement, resulting in reduced thermal energy in the preheater exhaust31. Annexes A and B provide guidance on best practices for accurately estimating waste heat availability and WHR system sizing. Project IRRs are a Function of WHR Utilization and Investment Costs All six of the plants that had installed WHR identified saving costs and improving efficiency as the primary drivers for installing the systems. Secondary reasons for implementing WHR included environmental considerations and implementing industry best practices to remain economically competitive. In fact, even with the reduced WHR utilization rates, the experience of the six plants shows that average electricity costs can be reduced by FIGURE 16. PROJECT IRRS ARE A FUNCTION OF 1.4 to 2.7 $/ton of clinker, contributing to better EBITDA INVESTMENT COSTS per ton of clinker produced. As shown in Figure 16, estimated project IRRs of the six installations ranged from 27% 25% 6 to 27 percent, driven largely by significant differences in investment costs. The range of WHR investment costs in terms of $/kW identified in the figure reflect the wide range 16% of reported total investment for the plants. While total 11% installed costs for WHR systems are generally a function 8% 30 Availability is defined as the ability of the WHR system to operate 6% during the time period it is expected to operate (that is, when the clinker production line is operating). 31 White cement requires raw materials with much lower iron and alumina content resulting in low melt liquid flow. Melt liquid acts as a solvent in final reactions in the kiln. Lower solvent flow results in longer residence 3,782 3,131 2,531 2,589 1,551 1,412 time in the kiln, requiring higher fuel consumption per ton of clinker. This is coupled with higher radiant heat loss in the kiln due to reduced coatings forming on the kiln refractories, resulting in less heat available 1-2 0-1 2-3 1-2 3+ 3+ for recovery in white cement production. WASTE HEAT RECOVERY IN THE TURKISH CEMENT INDUSTRY: CURRENT STATUS AND PROJECT EXPERIENCE 23 plants. The cost decrease is a combination of lower costs operations and maintenance concerns, and issues with from the system supplier and local scope providers, and training staff to operate the WHR system. While concerns recent adjustments to currency valuations. about the impact of WHR on kiln operations – kiln stability, clinker quality, production rate, and fuel consumption – Implementing WHR Projects May Require Dealing with are often cited as barriers to project implementation, none Issues Unfamiliar to Cement Plants of the surveyed plants reported any significant operational issues in these areas. The WHR installations all saved costs and generated reasonable returns on investment; however, the project Specific project development and operational issues development process itself was not without issues. As commonly cited include: shown in Figure 18, responses from the facilities all indicate some difficulties with project implementation ranging from • System design issues: Positive project economics problems with suppliers not meeting project milestones on depends on appropriate WHR system sizing and time, difficulties in the construction and commissioning of design. This requires accurate measurement of key the system, integration issues with the cement process, baseline process parameters such as preheater and clinker exhaust flows and temperatures, thorough understanding of expected kiln utilization FIGURE 17. WHR INVESTMENT COSTS and the moisture content of raw material and fuel, and consideration of potential variations in these parameters over time. Positive project economics also depends on quality system components and equipment installation. • Water conditioning: Required for both boiler feedwater and condenser cooling water, surveyed plants highlighted the need to utilize high-quality water conditioning chemicals to avoid erosion and corrosion problems, and to contract with reputable suppliers that not only provide quality chemicals, but that provide a full-service package that includes such tasks as daily tracking of conditioning equipment, water quality lab tests, online monitoring with analyzers, and corrosion measurements. • Training: Proper training of plant personnel on WHR FIGURE 18. PERCEIVED DIFFICULTIES WITH EXISTING WHR INSTALLATIONS system operation and maintenance as well as system performance monitoring was a critical need identified by multiple survey participants. WHR system O&M is unlike normal cement plant operations and may require new personnel with different technical skills and salary requirements. • Communication: The existing WHR systems were all installed by project teams led by foreign equipment vendors and engineering firms, with support from local companies for engineering support and installation. Communication can become complicated with multinational project teams such as this, particularly when there are language issues. 24 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY Implementation of Operational Best Practices Can of waste heat available for power generation. While the Enhance WHR Performance moisture content of delivered raw material may be beyond the control of plant operators, applying an understanding While the plants had positive returns on their WHR of the historical range and seasonal occurrence of investments and received substantial benefits in terms of variations in WHR system sizing and operational planning lower production costs, implementation of best practices will ensure optimum utilization of the equipment. Historical to better control factors that affect WHR performance data on as-received moisture content could potentially be could further enhance their bottom lines. One example used to reduce seasonal variations by planned blending of is management of the moisture content of raw materials. different sources. Applying proper storage and moisture Cement raw materials are received with an initial moisture management controls for materials stored on-site prior content varying from 1 to more than 50 percent, depending to grinding could further mitigate moisture issues and on source and season. Moisture is usually reduced to less promote optimum WHR performance. Annex B contains a than 1 percent before entering the kiln by use of preheater summary of best practices for both design and operation exhaust (and in some plants, clinker cooler exhaust) during of WHR systems. grinding. Variations in the amount of exhaust gas used for drying affects WHR performance by limiting the amount WASTE HEAT RECOVERY IN THE TURKISH CEMENT INDUSTRY: CURRENT STATUS AND PROJECT EXPERIENCE 25 3. There Is Potential for Additional WHR in the Turkish Cement Industry The survey gathered data on annual clinker design ORC WHR Is Economically Viable in the capacity and production as well as on preheater and Turkish Cement Industry cooler exhaust conditions from six plants currently without WHR. Based on this information, and using baseline A Base Case technical feasibility and financial analysis assumptions on WHR performance (see Annexes A and was conducted for each plant based on their reported C), a high-level feasibility analysis was conducted on each operating data for 2015, including information on kiln plant to estimate the potential heat available for recovery, utilization and operation, exhaust temperatures and flows and the size, cost, and baseline performance of applicable from the preheaters and clinker coolers, and the amount WHR systems. As a group, these six plants were smaller of heat used for drying raw material and fuel. ORC WHR than the plants surveyed with existing WHR systems, systems were sized for each plant based on the preheater ranging from 0.435 to 1.88 Mt/yr clinker design capacities and clinker cooler exhaust conditions, and the technical compared to 0.612 to 4.45 Mt/yr for the plants with WHR. and financial feasibility was estimated based on typical Only three of the plants had clinker capacities greater cost and performance assumptions for ORC systems than 1.0 Mt/yr, compared to five of the six plants with WHR, applied to cement plants. The basic assumptions used for and two of these plants with WHR had design capacities the feasibility analysis are detailed in Annex C. Chapter greater than 4.0 Mt/yr. 4 includes a sensitivity analysis on key factors impacting project economics, including WHR system utilization rates, Given the relatively small size of the plants in the group kiln operating hours, total investment cost, grid electricity without WHR, the relatively higher raw material moisture prices, currency exchange rates, WHR O&M costs, and content for these plants (Figure 6), the use of clinker WHR auxiliary power requirements. cooler exhaust for raw material drying by five of the plants, and the reported preheater and cooler exhaust The Base Case scenario is based on the maximum temperatures of 285oC to 350oC for this group (the six planned kiln operating hours in a given year as reported plants had a combined total of seven clinker lines: four by the plants, and does not account for potential variability with five preheater stages and three with four preheater in the amount of heat available throughout the year due to stages), the feasibility analysis was based on application of variations in moisture content or kiln operation. The model organic Rankine cycle (ORC) WHR. As discussed in Annex further assumes that all heat at the exit of preheater and A, there is overlap between application of SRC and ORC cooler is available for WHR, and the design is based on a systems when the exhaust temperatures are between 300 highly efficient integration with the plant, absence of leaks, and 350°C and WHR system capacities are between 4 and and the utilization of exhaust heat for drying materials only 10 MW. While selection of the appropriate WHR system in after the WHR, which may not always be the case. Table 4 this range can be affected by a variety of site conditions presents the key plant parameters, estimated WHR design and requires a direct comparison of cost, performance, sizing, and financial modeling results of the Base Case and benefits based on expected plant operation, ORC analysis. The ORC design capacities are estimated based technology was selected as the appropriate option for the on average temperature and flowrate data provided by feasibility analysis given that five of the six plants would the companies, although for appropriate sizing a precise utilize systems below 6 MW. metering of the exhaust streams for an extended period is needed. 26 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY TABLE 4. BASE CASE RESULTS Clinker Maximum Preheater Cooler WHR IRR, % Payback*, Cost of Kiln Kiln Exhaust Exhaust Capacity, years Generated Capacity, Operation, Used for Used for kW Electricity**, Mt/yr hrs/yr Drying? Drying? USD/MWh Plant 7 >1.0 < 2.0 8,300 Yes Yes 5,460 24% 4.8 20.3 Plant 8 <1.0 7,920 Yes Yes 2,150 12% 8.7 33.8 Plant 9 <1.0 8,592 Yes No 4,150 19% 6.1 25.2 Plant 10 1.0 to 2.0 8,000 Yes Yes 5,790 23% 5.0 21.0 Plant 11 <1.0 8,000 Yes No 4,230 17% 6.6 27.1 Plant 12 1.0 to 2.0 8,000 Yes Yes 6,000 23% 5.0 21.0 * Simple Paybacks based on a five-year average (2016–2020) of displaced electricity costs. ** Cost of Generated Electricity including both operating expenses and the 20 year Net Present Value of capital expense. Figure 19 shows the Base Case WHR design capacities are range from 2,500 USD/kW for the largest system of 6.0 within the international practice standard of 25 to 45 kWh/t MW to 3,800 USD/kW for the smallest system of 2.2 MW. clinker. In the Base Case, WHR systems provide from 13 to The estimated cost of electricity produced from the WHR 58 percent of the total electricity use depending on WHR systems varies between 21 and 34 USD/MWh compared to system size and the extent of finish grinding at each plant, an assumed cost of grid electricity of 78 USD/MWh32. This resulting in an estimated savings of 11 to 50 percent of results in a reduction in electricity costs of 1.3 to 8.0 USD/t annual electricity charges. (The higher-end estimates on clinker. (The range is $1.30 to $2.60 per ton of clinker percent of total electricity use and percent of electric bill when outlier Plant 11 is excluded.) savings are from Plant 11 in Figure 19, which is the outlier in the survey data provided. Without Plant 11 the higher Figure 20 shows the estimated simple payback periods 33 ends are 30% and 27%, respectively). While ORCs are only from the analysis as a function of WHR design capacity. beginning to be applied in the cement industry, they are As shown, the payback periods are driven largely by extensively used worldwide in geothermal applications the range in the total installed costs of the system used and in biomass heat recovery in Europe, and their capital in the analysis, which is a strong function of system size. and maintenance costs are relatively well understood. Payback periods range from 4.8 years for a 5.5-MW system Total investment costs used in the feasibility analysis to just under nine years for the smallest system (2.2 MW). Payback periods for systems in the 4 to 6 MW range were estimated at 5 to 7 years. Estimated IRRs ranged from 12 FIGURE 19 – ESTIMATED WHR DESIGN CAPACITY VS percent for the smallest system to 24 percent for systems CLINKER DESIGN CAPACITY larger than 5.5 MW as shown in Figure 21. The Turkish Cement Industry Still Has 125 to 230 MW of Feasible WHR Potential By the end of 2017, the Turkish cement industry has 54 active integrated cement plants with a total of 83.6 Mt/ yr of clinker production capacity. An additional three 32 The cost of electricity values of 21 to 34 USD/MWh include operating costs and the 20-year net present value of capex expenses. The Base Case estimated costs of electricity based on WHR operating costs only range from 7 to 17 USD/MWh. 33 Simple paybacks are based on a five-year average (2016 – 2020) of displaced electricity costs THERE IS POTENTIAL FOR ADDITIONAL WHR IN THE TURKISH CEMENT INDUSTRY 27 FIGURE 20. ESTIMATED PAYBACK FIGURE 21. ESTIMATED PROJECT IRRS integrated plants are under construction with total 5.5 Mt/ industry. Applying this standard to the 39 clinker plants yr of clinker capacity, and an existing plant is expanding that have not implanted WHR results in an estimate of with an additional 1.0 Mt/yr of clinker capacity. Average additional WHR potential ranging from 158 to 283 MW, with clinker capacity of the 54 plants is 1.55 Mt/yr. The smallest the corresponding range of investment potential totalling plant has a clinker capacity of 0.43 Mt/yr, with the largest $450 million to $630 million, as shown in Table 5 below. at 4.45 Mt/yr, and 15 of the plants have capacities below 1.0 Mt/yr. Fourteen of the plants ranging in size from The largest potential market, in terms of total remaining 0.68 to 4.45 Mt/y have already installed, or are in the WHR capacity of 88 to 158 MW (250 to 350 million USD), process of installing, a total 130 MW of steam-cycle WHR is in the 20 plants with clinker capacities of 1.0 to 1.99 Mt/ capacity, representing 30.6 Mt/yr, or 40 percent, of total yr. These plants could support WHR systems of 5 to 7 MW existing clinker capacity. While successful application of capacity. While, as noted throughout this report, project WHR depends on a variety of site-specific parameters, economics are driven by a number of site-specific factor the review of existing WHR systems and the feasibility such as number of preheater stages, raw material drying analysis of plants currently without WHR indicate that WHR requirements and number of individual kiln lines, systems is technically, and potentially economically, applicable to of this size had estimated simple payback periods of plants with more than 1 Mt/yr clinker capacity, adding up to around 5 years in the feasibility analysis. Both SRC and 125 to 250 MW of remaining WHR potential (25 plants, 43 ORC WHR systems could be applied in this size range Mt/yr clinker capacity). depending on plant specific parameters. The estimates of potential investment in this capacity range are based The analysis also verified that the existing industry on a mix of 50 percent ORC and 50 percent SRC capital standard of WHR producing between 25 and 45 kWh/t costs. The five plants with clinker capacities greater than of clinker is broadly applicable to the Turkish cement 2.0 Mt/yr represent a potential market of 41 to 73 MW (90 TABLE 5. REMAINING TECHNICAL POTENTIAL FOR WHR IN TURKISH CEMENT INDUSTRY Plant Clinker Number of Total Clinker Potential Potential Potential Potential Capacity, Mt/yr Plants Capacity Capacity Investment Capacity Investment (Mt/yr) @ 25 kWh/t @ 25 kWh/t @ 45 kWh/t @ 45 kWh/t (MW) (Million $) (MW) (Million $) <0.99 14 9.88 29.4 MW $111.5 52.9 MW $167.2 1.0 – 1.99 20 29.41 87.5 MW $247.7 157.6 MW $349.0 >2.0 5 13.60 40.5 MW $89.7 72.9 MW $116.6 39 52.89 157.4 MW $448.8 283.3 MW $632.8 28 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY to 115 million USD). This is a particularly significant market size would have expected payback periods of 4 years or because of the more compelling project economics of less due to economies of scale and the ability to utilize larger WHR systems. These plants could support WHR SRC technology. The estimates of potential investment for systems in a range of 7 to 15 MWs. WHR systems of this this capacity range are based on SRC system costs. THERE IS POTENTIAL FOR ADDITIONAL WHR IN THE TURKISH CEMENT INDUSTRY 29 4. Key Factors Impacting WHR Feasibility WHR Project Economics Are Driven by these factors drive the amount of electricity produced by the WHR system. Cost factors such as WHR capital Factors that Impact the Amount of WHR expenses, displaced electricity prices, and variations in Power Generated and Affect Project currency exchange rates also have significant impact Costs on project financial performance. Reasonable variations As described in Chapter 3, a technical feasibility and in WHR operating and maintenance costs and in WHR financial analysis was conducted for each plant, with Base auxiliary power requirements have more modest impact Case performance based on reported operating data for on project performance. 2015. ORC WHR systems were sized for the analysis based on preheater and clinker cooler exhaust conditions, and Specific details on each of the sensitivity factors are financial feasibility was estimated based on typical cost presented below in order of greatest to least impact on and performance assumptions for ORC systems. A series project financial performance: of analyses were then performed to assess the financial sensitivity of WHR projects to variations in key plant and WHR Capacity Utilization Is a Key system operating and financial parameters, as described Determinant of Power Generated and in Table 6. Detailed assumptions used in the sensitivity Project Economics analysis are presented in Annex C. Variations in WHR capacity utilization have the most Figure 22 presents the results of the sensitivity analysis significant impact on WHR project economics, as Figure 23 on project paybacks for a single plant (Plant 10), but are shows. WHR system capacity utilization can be affected by representative of the results for the entire survey group. a number of factors including: (i) changes in kiln operations, As shown, WHR capacity utilization and kiln operating (ii) variations in raw material and fuel moisture that reduce hours have the greatest impact on project economics as available heat to the WHR system, (iii) extended downtime on the WHR system itself due to operational issues, (iv) oversizing the WHR system in the initial system design, TABLE 6. DESCRIPTION OF FINANCIAL ANALYSIS (v) less production in different terms depending on sales. SCENARIOS FOR ORC WHR Scenario Description FIGURE 22. SENSITIVITY ANALYSIS ON WHR PROJECT Base Case Operations as per maximum planned operating PAYBACKS hours of cement plant in given year at 100% WHR capacity utilization, with heat used for drying limiting available heat for WHR as per typical drying practice and as reported by each plant. 1 WHR Capacity Utilization: -30% / +5% 2 Clinker Kiln Operating Hours: 6034 hours, 8595 hours 3 Total Capital Cost: -10% / +20% 4 Electricity Prices: +10% / -10% 5 TL/$ Exchange Rate: -10% / +10% 6 WHR O&M Expenses: -130% / +30% 7 WHR Auxiliary Power: 8%, 10% (base case), 12% 30 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY A WHR capacity utilization of 70 percent (the Base Case FIGURE 23. IMPACT OF WHR CAPACITY UTILIZATION is 100 percent) increases paybacks by 5.3 years for the RATES ON PROJECT PAYBACK smallest WHR system and 2.4 years for systems greater than 5 MW. A WHR capacity utilization of 105 percent, which could represent increased kiln operating hours or additional waste heat availability, improves Base Case paybacks by 0.25 to 0.5 years. Kiln Operating Hours also Drive WHR Project Performance Similar to WHR utilization, kiln operating hours also have a significant impact on WHR project economics, as shown in Figure 24. The Base Case kiln operating hours were the maximum operating hours reported by each plant and ranged from 7,902 to 8,592 hours per year. Applying 6,034 annual kiln operating hours to each plant (this is FIGURE 24. IMPACT OF KILN OPERATING HOURS ON the lowest actual operating hours for 2015 reported in the PROJECT PAYBACK survey) increases paybacks by 3.7 years for the smallest WHR system and by 1.6 years for the largest. Conversely, applying 8,595 annual kiln operating hours (close to the maximum annual hours achievable, considering required downtime for annual maintenance on the kiln) shows a decrease in payback by up to 0.8 years. Not surprisingly, WHR Capital Cost (CapEx) Has Significant Impact on Project Economics Variations in the total investment cost of WHR systems also have a substantial impact on project economics, and differences in actual project capital costs from industry averages are commonplace due site-specific conditions FIGURE 25 – IMPACT OF CAPEX ON PROJECT and site limitations that can subtract, or more often, add PAYBACK to, projected investment costs. Figure 25 shows the effect of variations in total project capital costs (equipment and installation/construction costs) on financial performance. A 10 percent decrease in total investment costs from the Base Case decreases paybacks from 0.5 to 1.0 years. A 20 percent increase in total investment cost increases paybacks by 2.3 years for the smallest WHR system and 1.1 year for the largest. It should be noted that capex can vary significantly depending on the technology provider used, which may have a direct impact on the payback periods and overall feasibility of the project. KEY FACTORS IMPACTING WHR FEASIBILITY 31 Displaced Electricity Prices Drive Project at the level of yearly average wholesale market prices in Turkey during that year, including fees for transmission, Savings distribution, and renewable energy support. The Base Figure 26 shows the sensitivity of project paybacks to a +/- Case assumes a 2018 price for electricity to industrial 10 percent variation in electricity prices over the Base Case users of 0.289 TL/kWh (0.078 USD/kWh at the Base Case projection. A 10 percent decrease in price results in an exchange rate). The rate of increase for the out-years increase in project paybacks by 0.62 years for the largest is based on forecasts developed in Mercados Market system and 1.25 for the smallest. A 10 percent increase Report, March 2017. Mercados’ price forecast is influenced in price results in a decrease in project paybacks ranging by assumptions on fuel price outlook, planned capacity from 0.49 years for the largest system to 0.97 for the additions, attractiveness of new investments, future carbon smallest. WHR provides costs savings for cement plants prices, growing demand, decreasing efficiency gains in by replacing electricity purchases from the grid with an new thermal power plants over time, and decreasing costs upfront capital investment and modest ongoing O&M costs of renewable power. The forecast projects substantial for the WHR system. As such, the price of grid electricity, price increases in the near term (through 2020) due to and the price outlook into the future, has significant impact rising oil prices and depreciation of existing natural gas on net savings and WHR project economics. The cost of power plants. Current over-capacity ends in 2021 as a electricity paid by cement plants in each year was assumed tighter demand-supply balance is reached due to strong market fundamentals, and new capacity is needed. The forecast projects the industrial electricity rate to increase FIGURE 26. IMPACT OF ELECTRICITY PRICES ON PROJECT PAYBACK at an average annual rate of 9.4 percent between 2018 and 2023, reaching a price of 0.447 TL/kWh (0.103 $/kWh at the Base Case exchange rate) in 2023. Prices continue to increase more gradually from 2023 onward (average annual rate of 1.9 percent between 2024 and 2040) due to forecasts of gradually increasing global oil prices and gradually increasing carbon costs starting in 2025. Variations in Currency Exchange Rates Affect Project Costs Since the leading suppliers of ORC WHR systems are foreign entities, volatility in the currency exchange rate between the Turkish lira and U.S. dollar creates uncertainty in the financial evaluation of potential WHR projects. Empirical studies have also shown a negative FIGURE 27. IMPACT OF CURRENCY EXCHANGE RATES relation between exchange rate volatility and plant level ON PROJECT PAYBACK investment, particularly for industries with significant export exposure. The Turkish lira experienced a significant depreciation against hard currencies over the past two years. The exchange rate has ranged from 2.335 TL/$ in January 2015 to 3.779 TL/$ in February 2017. The Base Case assumption for the financial analysis is 3.698 TL/$ for 2018. Forecasted values of the exchange rate are based on estimates by the Economist Intelligence Unit, accessed on July 21, 2017. The forecast projects the exchange rate increasing at an average annual rate of 3.2 percent for the first five years of the forecast, and then ramping down over the long term to a 2.0 percent annual increase in 2040. Figure 27 shows the sensitivity of project paybacks to 32 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY a +/-10 percent variation in exchange rate over the Base Reasonable Variations in Auxiliary Power Case. A 10 percent increase in the exchange rate, to 4.068 Requirements Have Minimal Impact on TL/$, results in an increase in project paybacks from 0.54 years for the larger systems to 0.96 for the smallest. A 10 Project Performance percent decrease in the exchange rate, to 3.328 TL/$, WHR systems have significant auxiliary power results in a decrease in project paybacks ranging from requirements to operate the controls, pumps, blowers, and 0.43 years for the larger system to 0.79 for the smallest. cooling towers necessary for system operation. Auxiliary power requirements typically range from 8 to 12 percent of Variations in O&M Costs (Opex) Have gross generation depending on system configuration and Only a Modest Impact on Financial whether condensers are air cooled or water cooled. The Base Case assumption was 10 percent. Figure 29 shows Performance the relatively modest impact of variations in auxiliary Because WHR generates power by recovering heat power consumption of 8 percent and 12 percent of gross normally wasted in the cement plant and does not rely on generation, with paybacks decreasing or increasing from any additional fuel consumption, the only costs for operating 0.1 to 0.25 years depending on WHR system capacities. the WHR system are the operating and maintenance (O&M) costs related to ongoing system maintenance, day- Subsidies and Financing Mechanisms to-day monitoring and control, and servicing. WHR O&M May Improve Feasibility for WHR includes daily, monthly, and annual equipment inspection and adjustments, periodic preventive maintenance and Investments refurbishment, and restoration and repair from unplanned All the financial analysis in the report has been done equipment failures, as well as daily control and monitoring assuming no incentives are used in the project cycle. of system performance. O&M costs can vary based on However, there are many incentives that can be whether services are provided by a third party or by in- considered which may substantially improve the feasibility house staff. The Base Case assumed annual O&M costs of a WHR investment. In this section a specific incentive at 3 percent of system investment cost for systems of 2-3 has been analyzed in detail. Further information regarding MW capacity, 2.5 percent for systems of 3-5 MW capacity, other incentives and financing mechanisms is provided in and 2.0 percent for systems of > 5 MW. Figure 28 shows Annex C. the relatively modest impact of +/- 30 percent variations in these Base Case levels, with paybacks decreasing On May 9, 2014, the 2014/6058 Decision in the 28995 or increasing from 0.15 to 0.7 years depending on WHR Number Official Gazette further clarified that energy system capacities. efficiency and waste heat recovery investments are added FIGURE 28. IMPACT OF O&M COSTS ON PROJECT FIGURE 29. IMPACT OF AUXILIARY POWER PAYBACK REQUIREMENTS ON PROJECT PAYBACK KEY FACTORS IMPACTING WHR FEASIBILITY 33 TABLE 7. INCENTIVES AND STATE GRANTS IN INVESTMENTS SCHEME Regional Investment Incentives Scheme Instruments Incentive Instruments Region I II III IV V VI VAT Exemption YES Customs Duty Exemption YES Tax Reduction Rate (%) 50 55 60 70 80 90 Tax Reduction Reduced Tax Rate (%) 10 9 8 6 4 2 Rate of Out of OIZ* 15 20 25 30 40 50 Contribution to Within OIZ* 20 25 30 40 50 55 Investment (%) Social Security Support Out of OIZ* 2 years 3 years 5 years 6 years 7 years 10 years Premium Period Within OIZ* 3 years 5 years 6 years 7 years 10 years 12 years Support (Employer’s Upper Out of OIZ* 10 15 20 25 35 No limit Share Limit for Within OIZ* 15 20 25 35 No limit No limit Support (%) Land Allocation YES Interest Rate TRY Denominated Loans (points) N/A N/A 3 points 4 points 5 points 7 points Support FX Loans (points) 1 point 1 point 2 points 2 points Social Security N/A N/A N/A N/A N/A 10 years Premium Support (Employee’s Share) Income Tax Withholding Allowance N/A N/A N/A N/A N/A 10 years Source: Investment Support and Promotion Agency Turkey. * OIZ: Organized Industrial Zone. to the top priority list, and all priority investments will be As shown in the example above, for a theoretical 40 supported with “5th region” incentives (independent of million TL WHR investment and total investment incentive location) under the Ministry of Economy’s “Incentives and certificate period of two years, a cement company can State Grants in Investments Scheme.” Under this scheme, save up to 16 million TL on taxes, with the exact amount the investor does not pay value-added tax (VAT) for the depending on the company’s income. The maximum WHR equipment; the investor will not pay customs tax for benefit limit here is the cap of 40 percent of investment imported equipment; the social security premiums for WHR (or 50 percent if plant is located in an OIZ). The same plant operating staff is supported by the ministry; the ministry can at the same time benefit from interest rate support if will cover five percentage points of bank loan interest in the loan is taken for investment and other relevant items, the first year; and the tax is reduced during the investment as presented in Table 7. period by up to 40 percent of the total WHR investment (and by up to 50 percent in an organized industrial zone). Example (1): Interest rate support Example (2): Tax reduction Principal = 1 million TL Total WHR Investment = 40 million TL Interest Rate = 15% Assuming company EBITDA of 50 million TL First Payment is 180 days later And annual tax (20%) payment of 10 million TL Interest = (Principal * Int. Rate * Days) / (days in year) Incentives will lead to a reduced tax of (4%) = 2,000,000 TL = 1,000,000 * 0.15 * 180 / 360 The difference of 8 million TL is called deferred tax and it will = 75,000 TL is the interest cost not be paid until reaching the amount of contribution cost With incentive: (40% of investment cost), which makes 16 million TL total, = 1,000,000 * 0.05 * 180 / 360 during investment period. = 25,000 TL will be paid by ministry 34 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY Annex A – Waste Heat Recovery Technologies Introduction • What is the load factor of the waste heat source – are the annual operating hours sufficient to amortize the Waste heat recovery for power (WHR) is the process of capital costs of the WHR system? capturing heat discarded by an industrial process or • Does the temperature of the waste stream vary over prime mover and using that heat to generate electricity. time? Energy-intensive industrial processes – such as those in • What is the flow rate of the waste stream and does it refineries, steel mills, glass furnaces, and cement kilns – vary? all release hot waste streams that can be harnessed with • Is the waste stream at a positive or negative pressure, well-established technologies to generate power. To be and does this vary? effective, WHR must have a source of waste heat that is of • What is the composition of the waste stream? sufficiently high temperature for the waste heat recovery • Are there contaminants that may corrode or erode system to be both thermodynamically and economically the heat recovery equipment? feasible. In addition, the best sources of waste heat for • Is there space available in or close to the waste heat bottoming cycles are high-volume and high-load factor so stream for recovery and generation equipment? that the power generation equipment can operate with economies of scale and the capital cost can be offset The answers to these questions affect technology choice, by nearly constant output throughout the year. The key system design and, ultimately, the economic viability of advantage of waste heat recovery systems is that they the waste heat recovery application. This Annex describes utilize heat that would otherwise be wasted from an the basic requirements of heat engine systems to produce existing thermal process to produce power, as opposed power in the context of utilizing industrial waste heat to directly consuming additional fuel, displacing higher- streams, and provides details on two commonly used priced purchased electricity for the user, and reducing technologies for producing power from waste heat. overall energy use and emissions. At the project level, a number of factors in addition to the temperature of the waste heat must be considered to determine the feasibility of power generation from Rankine Cycle waste heat sources. While many high-temperature waste Waste heat recovery power systems used for cement heat sources are straightforward to capture and use kilns operate on the Rankine Cycle34. In a Rankine cycle, with existing technologies, other sources are filled with heat is supplied externally to a fluid in a closed loop. contaminants and can be expensive to recover because Water is the most commonly used fluid – steam turbines the waste streams must be cleaned prior to use. Not only found in thermal power generation plants produce most can the cleaning process be expensive, but removing of the electricity in the world, including power from coal, contaminants prior to use often removes heat at the same biomass, solar thermal, and nuclear energy. However, time. Other waste heat sources are difficult to recover systems using other fluids or combinations of fluids have because of equipment configuration or operational issues. been developed that have advantages over water in Along with temperature, a project developer would need certain WHR applications. to consider a number of questions about the candidate waste heat source: In a heat recovery Rankine cycle, a working fluid in the liquid state is first pumped to elevated pressure before • What is the availability of the waste heat – is it entering a heat recovery boiler (as illustrated in Figure continuous, cyclical, or intermittent? 34 The Rankine cycle, named after William John Macquorn Rankine, is a thermodynamic cycle that converts heat into mechanical work. ANNEX A – WASTE HEAT RECOVERY TECHNOLOGIES 35 is then exhausted to a condenser at vacuum conditions FIGURE A-1. RANKINE CYCLE HEAT ENGINE where heat is removed by condensing the vapor back into a liquid. The condensate from the condenser is then returned to the pump for continuation of the cycle. By condensing the working fluid to a liquid, the work required by the pump consumes only 1 to 3 percent of the turbine power and contributes to a higher cycle efficiency35. However, the range between the heat source and heat sink temperatures is typically much lower than for an open cycle combustion turbine, limiting both theoretical and practical efficiencies of the Rankine cycle. In addition, as a result of irreversibility in various components such as fluid friction and heat loss to the surroundings, the compression by the pump and the expansion in the turbine are not isentropic, somewhat increasing the power required by the pump and decreasing the power generated by the turbine, and the actual cycle efficiency deviates from the ideal Rankine cycle efficiency. For WHR applications, the achievable Rankine cycle efficiency typically ranges from 30 to 50 percent of the Carnot efficiency. Commercially available waste heat recovery Rankine cycles are based on steam or an organic compound as the working fluid36. Steam Rankine Cycle (SRC) Steam Rankine cycles are the most common waste heat recovery systems in operation in cement plants and have the following attributes: 35 One of the principal advantages of the Rankine cycle compared to other thermodynamic cycles is that during the compression stage relatively little work is required to drive the pump, since the working fluid is in the liquid (nearly incompressible) phase. The energy required for raising the pressure of the fluid is a function of the change in volume; A-1). The pressurized fluid is vaporized by the hot exhaust, therefore, raising the pressure of a compressible gas as is required in a combustion turbine (Brayton Cycle) requires much more energy than and then expanded to lower temperature and pressure raising the pressure of an incompressible liquid. 36 in a turbine, generating mechanical power that can drive A third type of Rankine cycle is based on a mixture of water and ammonia and is called the Kalina cycle. Application of the Kalina cycle to an electric generator. The low-pressure working fluid the cement industry is still in the demonstration phase. TABLE A-1. STAGES OF THE RANKINE CYCLE 1-2: Isentropic compression (pumping) The working fluid enters the pump as saturated liquid (state 1) and is compressed isentropically to the operating pressure of the boiler (state 2) 2-3: Constant pressure heat addition Saturated liquid (state 2) enters the boiler and leaves it as superheated vapor (state 3) 3-4: Isentropic expansion (turbine generator) Superheated vapor expands isentropically in a turbine and produces work, and exits the turbine as low-pressure, lower-temperature saturated vapor (state 3 to state 4) 4-1: Constant pressure heat rejection Low-pressure steam (state 4) is condensed in the condenser to a saturated liquid (state 1) 36 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY • Most familiar to the cement industry and are if compared to water, organic fluids do not need to be generally economically preferable where source superheated to avoid condensation in the turbine during heat temperature exceeds 350oC. expanding stage; moreover, the organic fluids have higher • Based on proven technologies and generally simple molecular weight, lower boiling point, and higher vapor to operate. pressure than water. All these features lead to some key • Widely available from a variety of suppliers. technical advantages compared to conventional steam • Generally have lower installation costs than other cycles in certain applications: Rankine cycle systems on a specific cost basis ($/kW). • Need higher-temperature waste heat to operate • Higher turbine efficiency. optimally (minimum 260oC); generation efficiencies • Low mechanical stress on the turbine (low tip speed, fall significantly at lower temperatures, and steam moderate temperature). conditions with lower pressure and temperature • No blade erosion (no liquid particles during can result in partially condensed steam exiting the expansion, due to the shape of saturation curve). turbine, causing blade erosion. • No oxidation (some organic fluids can even be • Usually require a full-time operator, depending on considered as lubricants themselves). local regulations. • Higher efficiency with low and moderate temperature • Require feedwater conditioning systems. sources (for example, 24% with a 300°C heat source). • Generally require a water-cooled condenser; air- cooled condensers can be used but create a ORC systems can be utilized for waste heat sources as performance penalty due to higher condenser low as 150oC, whereas steam systems are limited to vacuum pressures. heat sources greater than 260oC37. The ORC systems • In general, match well with large kilns and systems are typically designed with two heat transfer stages. with low raw-material water content (resulting in The first stage transfers heat from the waste gases to an higher exhaust gas temperatures). intermediate heat transfer fluid (such as thermal transfer oil). The second stage transfers heat from the intermediate Organic Rankine Cycle (ORC) heat transfer fluid to the organic working fluid. The ORCs have commonly been used to generate power in Organic compounds with better generation efficiencies at geothermal power plants, and more recently, in pipeline lower heat-source temperatures are used as the working compressor heat recovery applications. ORC systems fluid in organic Rankine cycle (ORC) systems. The most have been widely used to generate power from biomass widely used organic fluids are hydrocarbons (such as systems in Europe. A growing number of ORC systems pentane), siloxanes (employed also in cosmetic products), have been installed on cement kilns38. and refrigerants (more common in HVAC systems and refrigeration). In Figure A-2, saturation curves for several The actual configuration of ORC systems can be different organic fluids are shown: as the shape of curves suggest, depending on the specific application and site conditions, such as type of heat source, demand for low temperature FIGURE A-2. SATURATION CURVES FOR SEVERAL heat, availability of water, space constraints. Usually, when ORGANIC FLUIDS the primary heat sources are hot, and dusty combustion 37 Low-pressure, low-temperature steam produced from lower temperature heat sources will partially condense in the last stages of the steam turbine, resulting in low efficiency and potential blade erosion. 38 Two ORC systems began operating in the late 1990s in cement plants: a 1.2 MW system installed in 1999 at the Heidelberg Cement plant at Lengfurt, Germany, recovers heat from the clinker cooler vent air; the second ORC system is a 4.8 MW unit located at AP Cement (now Ultra Tech Cement), Tadipatri, Andhra Pradesh, India. ORC system (2 MW) at Italcementi’s Ait Baha plant in Morocco in 2010 (5,000 tpd clinker line). In 2012, 4 MW unit at a Holcim Romani plant in Alesd (4,000 tpd clinker line); Holcim Slovakia (5 MW at 3,600 tpd line at the Rohoznik plant) and an undisclosed North American plant (7 MW). Holcim is installing another 4.7 MW ORC system at its Mississauga, Canada plant from an undisclosed provider. Jura cement 2.0 MW ORC system at the Wildegg AG plant in Switzerland. Source: Turboden. ANNEX A – WASTE HEAT RECOVERY TECHNOLOGIES 37 gases and cooling water are available, the heat recovery transfer fluid (thermal oil, hot water, or steam), and then systems consist essentially of a primary heat exchanger expands in a turbine (4-5), which is usually directly coupled (the ORC unit) and a cooling system for dissipating heat of to the generator. Condensation (8-1) is performed by a condensation downstream (the ORC turbo-generator). As cooling medium (air or water) after the regenerator. The mentioned earlier, the primary heat exchanger transfers the cycle is closed when liquid from the condenser is pumped waste heat from the exhaust gas to the ORC unit by means (1-2) to reach the evaporation pressure. An internal heat of a heat carrier (typically thermal oil, pressurized water, or exchanger (the “regenerator”) is placed downstream of saturated steam). This is one of the main differences with the turbine in order to achieve higher efficiency (5-8, 2-7). conventional steam Rankine cycles – while the waste heat recovery boiler evaporates steam directly using the heat Specific features impacting ORCs in cement industry of the exhaust, in ORC systems the heat is transferred from applications include the following: the waste heat recovery units through a heat carrier, which is heated by the hot exhaust, and transfers the heat to the • Simple start-stop procedures, with quiet running as organic fluid in the closed-loop ORC. well as automatic and unattended operation. • High availability (about 98%). The ORC unit converts the incoming thermal energy into • High flexibility (good efficiency at partial load, with electricity and heat at relatively low temperature. The heat possible operation as low as 10% or less of nominal discharged from the power cycle during condensation power). is then released to the environment by means of an • Long life and reduced O&M requirements. intermediate water circuit (or mixture of water and glycol • Can recover heat from gases at lower temperatures to prevent freezing in winter). The dissipation of this than is possible with conventional steam systems, heat can be in the form of a dedicated system: this can enabling ORCs to utilize all recoverable heat from the be either a dry system, with air-coolers, or a wet system air cooler. with evaporative cooling towers. As an alternative, direct • Operate with condensing systems above atmospheric condensation of the working fluid through an Air Cooled pressure, reducing risk of air leakage into the system Condenser (ACC) is often preferable, for both operational and eliminating the need for a de-aerator. and technical reasons – first, avoiding one heat transfer • Not susceptible to freezing. passage increases the overall efficiency of the system; • Because ORCs operate at relatively low pressure, and second, an ACC does not require any water supply they can operate unattended and fully automated in and/or water treatment. many locations depending on local regulations. • Can utilize air-cooled condensers without negatively Figure A-3 shows a simplified scheme for a typical waste impacting performance. heat recovery ORC system. Following the cycle schematic, • Lower-speed (rpm) ORC turbine allows generator the working fluid is first pre-heated (7-3) and evaporates direct drive without the need for and inefficiency of a (3-4) by means of the heat carried in the intermediate heat reduction gear. • ORC equipment (turbines, piping, condensers, heat exchanger surface) is typically smaller than that FIGURE A-3. TYPICAL ORC SCHEME AND required for steam systems, and the turbine generally THERMODYNAMIC CYCLE consists of fewer stages. • Although ORCs can provide generation efficiencies comparable to a steam Rankine system, ORCs are typically applied to lower temperature exhaust streams, and limited in sizing and scalability, and generally are smaller in capacity than steam systems. • Depending on the application, ORC systems often have a higher specific cost ($/kW) than steam systems • The two-stage heat transfer process creates some system inefficiencies. 38 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY • The heat transfer fluids and organic fluids normally water, kiln operation, raw-material moisture and resulting used in ORCs are combustible, requiring fire drying needs, and system capital costs. In addition, any protection measures and periodic replacement over anticipated changes to the cement production process time. Also, there may be environmental concerns should be considered, as they may affect the potential over potential system leaks. and performance of any installed WHR system. Relevant • In general, ORC systems are well-matched with small- operational changes to be considered in advance to medium-size, high-efficiency kilns or kilns with include, for example, any planned increase or decrease elevated raw-material moisture content (resulting of production capacity over the long term; increase or in lower waste gas temperatures due to increased modification of pre-heater stages; or the introduction or preheating or raw material drying). increase in use of alternative fuels (for example, refused derived fuel). WHR Technology Selection Comparative advantages and disadvantages of SRC and The amount of heat available for waste heat recovery, ORC systems that could impact technology choice for in combination with the temperature at which the heat cement applications are summarized in Table A-2: is available, are the main drivers for the choice between SRC and ORC technologies. In general, ORC technologies Figure A-4 is a simple decision tree that presents guidelines are often a preferred choice for projects with up to 5 for selecting the appropriate WHR system based on key MW of power output and when heat is rejected at lower threshold questions, including water availability at the temperatures (less than 300°C), while steam cycles are plant, amount of waste heat available in the preheater normally preferred for larger systems and when heat is and cooler exhaust streams, and the temperature of the available at higher temperature (above 350°C). However, exhaust streams. Note that there is some overlap between there are a variety of other considerations that can application of SRC and ORC systems when the exhaust further impact technology choice, such as availability of TABLE A-2. COMPARISON OF SRC AND ORC TECHNOLOGY COMPARISON BETWEEN ORC AND SRC CHARACTERISTICS TECHNOLOGY ADVANTAGES DISADVANTAGES • Applicable to medium-low temperature • Less efficient than SRC in medium-high applications (heat source <300°C ) temperature applications (heat source • Better choice for small installations (<2-5 MW >350°C ) power output) • Higher capital cost on a USD/kW basis • Very high turndown ratio (power production is • Higher auxiliary power requirements due to possible at ratios as low as 10% of nominal heat increased pump and fan power ORC input) • Organic fluids can be expensive and require more • Better choice if water is not easily and attention than water when handled (they may be abundantly available harmful to environment or flammable) • Unattended operation (no dedicated supervision • Leakage detection is a must, to avoid losses of or operators required) expensive fluid • Low turbine maintenance requirements • Emerging technology with limited installations worldwide in cement plants • Well-known and reliable technology, with many • Requires large amounts of water installations worldwide in cement plants • Requires expensive equipment and chemicals for • Lower USD/kW CapEx for systems of equal water treatment size; Chinese suppliers have quoted 4 to 6 MW • Less flexible and efficient at part load operation (at systems at 2,000 USD/MW installed input loads <60-70%) • More efficient than ORC in medium-high • Turbine more prone to maintenance requirements SRC temperature applications (heat source (for the effects of corrosion or erosion) >350°C ) • Better choice for large installations (>5 MW power output) • Water is normally easily available and environmentally friendly ANNEX A – WASTE HEAT RECOVERY TECHNOLOGIES 39 TABLE A-3. WHR SYSTEM & TECHNOLOGY SUPPLIERS STEAM SYSTEMS Japan Kawasaki Plant Systems Ltd. (JPN) (Dual Pressure Steam System) Anhui Conch/Kawasaki Engineering Co., Ltd. (CHN/JPN) (Dual Pressure Steam System) Sinoma Energy Conservation Ltd. (CHN) (Single Pressure Steam System) Nanjing Triumph Kenen Environment & Energy Co., Ltd. (CHN) Nanjing Triumph (Kesen) Dalian East New Energy Development Co., Ltd. (CHN) (Dual pressure steam system) CITIC Heavy Industries Co., Ltd. (CHN) China (Dual-pressure steam system) China National Building Materials Group (CNBM) (CHN) (Single-pressure steam system) Hefei Cement Research Design Institute (HCRDI) (CHN) Shanghai Triumph (Kesen) Energy Conservation (STEC) – JV between China Triumph International Eng. (CTIEC) and Mitsubishi (CHN/JPN) China National Building Materials Group (CNBM) (CHN) (Single-pressure steam system) Transparent Energy Systems Private Limited (IND) India Tecpro Systems Limited/NTK (IND/CHN) Thermax/Taiheyo Engineering (IND/JPN) FLSmidth (DEN) and others… ORGANIC RANKINE CYCLE SYSTEMS ORMAT (USA) Turboden / Mitsubishi (JPN) ABB (CHE) Exergy (ITA) Opcon Energy and others… KALINA CYCLE SYSTEMS Wasabi Energy (Australia) And others… 40 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY temperatures are between 300 and FIGURE A-4. COMPARISON OF SRC AND ORC TECHNOLOGY 350°C and the available waste heat is Simple decision-making tree for waste heat recovery projects between 20 and 100 MW. Selection of the appropriate system in this range can Steam Rankine Cycle Organic Rankine Cycle (SRC) UNCERTAIN (ORC) be affected by a variety of site conditions and requires a direct comparison of cost, Preliminary measurement performance, and benefits based on performed expected plant operation. + Reliable data available WHR System Suppliers The report maps out major WHR system Water scarcity YES suppliers, which are divided into three groups according to the technology. The detailed information about the suppliers NO can be seen in “Waste Heat Recovery for the Cement Sector” report published on June 2014, that can be found in IFC Waste Heat available YES website. > 100 MW NO Waste Heat available YES < 20 MW NO YES Exhaust Temperature > 350oC NO Exhaust Temperature YES < 300oC NO Further investigations required Direct comparisons of different technologies’ performances, costs, benefits ANNEX A – WASTE HEAT RECOVERY TECHNOLOGIES 41 Annex B – Best Practices in WHR Design and Operation Application of waste heat recovery power systems to • Number of preheater/precalciner stages. cement kilns can be challenging. The exhaust gases from • Configuration of the clinker cooler system. the kiln preheaters and clinker cooler typically contain • Moisture content of the raw-material feed (determines relatively high dust concentrations that sometimes exceed heat requirement for the kiln and the amount of 50 grams per nanometer cubed and the waste gas preheater exhaust needed for drying). temperatures can fluctuate widely during kiln operation. • Moisture content of the fuel to the kiln. Furthermore, many plants utilize a portion of the preheater • Efficiency of the top-stage cyclone (determines dust exhaust gas to dry raw materials, and the amount available content in the preheater exhaust). for heat recovery can vary widely depending on the • Amount of excess air in the kiln. moisture content of the raw feed39,40. • Amount of air infiltration. • Annual operating hours and capacity factor of the kiln. Evaluating Recoverable Waste Heat and It is critical to properly analyze each of these parameters in Power Generation Potential establishing a baseline for evaluating waste heat potential The amount of waste heat recoverable from a cement kiln and project feasibility. Care must be taken to ensure depends on the volume, temperature, and composition of that baseline assumptions reflect the actual operating the preheater exhaust and the air cooler exhaust. Exhaust conditions expected during annual operation. Seasonal flows and temperatures are, in turn, influenced by a or daily variations, or design changes, in any of these number of factors including: factors can significantly impact WHR performance and the amount of power produced, affecting project economics 39 Lawrence Berkeley National Laboratory (LBNL), “Energy Efficiency and expected paybacks. Improvement Opportunities for the Cement Industry, Worrell, Galitsky,” Price, January 2008. 40 “Desk Study on Waste Heat Recovery in the Indian Cement Industry,” Confederation of Indian Industry, Final Report, April 2009. TABLE B-1. TYPICAL AVAILABLE HEAT FROM PREHEATER FOR DRY PROCESS KILNS Parameter Unit Preheater kilns Preheater with precalciner (number of stages) Number of cyclone 4 4 5 6 stages Kiln capacity range TPD 1000 - 2500 2000 – 8000 Top-stage exit Deg C 400 340 300 260 temperature GJ / tonne clinker 0.904 0.771 0.678 0.586 (kcal/kg) (216) (184) (162) (140) Heat available in preheater exhaust GJ / hr for 1 MTPA* 113.0 96.4 84.7 73.3 (Mkcal/hr) (27.0) (23.0) (20.3) (17.5) Specific thermal energy GJ / tonne clinker 3.55 3.14 3.01 2.93 consumption (kcal/kg) (850) (750) (720) (700) Source: “Desk Study on Waste Heat Recovery in the Indian Cement Industry,” Confederation of Indian Industry, Final Report, April 2009. * MTPA – million metric tons per annum. 42 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY Preheater Stages Figure B-1 shows the power generation potential for a steam waste heat recovery system applied to the exhaust The number of preheater stages in a cement plant of a typical 5,000 tpd clinker line for exhaust temperatures has significant bearing on the overall thermal energy ranging from 300 to 450°C. consumption and waste heat recovery potential. The higher the number of stages, the higher the overall thermal energy efficiency of the kiln and the lower the Air Cooler Configuration potential for waste heat recovery. Selection of the number The clinker cooler design also impacts waste heat of preheater stages is based on several factors, such as availability. The basic cooler function is to remove heat cooler efficiency, restrictions on preheater tower height, or from hot clinker discharged from the kiln so that the clinker heat requirements for the mill itself. Table B-1 summarizes can be handled by subsequent equipment. Rapid cooling the efficiency (specific heat consumption) and quantity also improves clinker quality and grindability. Typically, of waste heat recoverable from state-of-the-art kilns. state-of-the-art coolers are grate coolers, which have Preheater exhaust temperatures range from 400°C for various stages of development. Table B-2 summarizes the small kilns with four preheater stages, to below 300°C for heat available in different generations of grate coolers. large kilns with six preheater stages. Exhaust air temperatures from the clinker cooler range from 250 to 340°C depending on cooler configuration and recuperation efficiency. Waste heat recovery potential FIGURE B-1. POWER GENERATION POTENTIAL AS A depends on the type and generation of cooler and the FUNCTION OF PREHEATER EXHAUST TEMPERATURE extent of utilization of cooler exhaust for the raw material or coal mills. Power Generation Potential Table B-3 shows the total heat available from both the preheater exhaust and clinker cooler air for a typical 5,000 tpd clinker plant. Power conversion efficiencies range from 18 to 25 percent, resulting in potential power capacities of 6 to 9 MW. Typically, the potential power generation, depending on waste heat losses and the number of preheater cyclone stages, ranges from 25 45 kWh/t of clinker. Assuming a Source: PENTA Engineering 2013. typical plant electrical power requirement of 106 kWh/t TABLE B-2 – TYPICAL AVAILABLE HEAT FOR GRATE CLINKER COOLERS Parameter Unit 1st Generation 2nd Generation 3rd Generation Grate Plate Type Vertical aeration with Horizontal aeration Horizontal aeration holes in plate Cooling Air Input Nm3/kg clinker 2.0 – 2.5 1.8 – 2.0 1.4 – 1.5 Exhaust Air Volume Nm3/kg clinker 1.0 – 1.5 0.9 – 1.2 0.7 – 0.9 GJ / Tonne clinker 0.419-0.502 0.335-0.419 0.293-0.335 (kcal/kg (100 – 120 (80 – 100) (70 – 80) Heat Available in Exhaust GJ / hr for 1 MTPA* 52.3-62.8 41.9-52.3 36.6-41.9 (Mkcal/hr) (12.5 – 15.0) (10.0 – 12.5) (8.8 – 10.0) Recuperation Efficiency % <65 <70 >73 Source: “Desk Study on Waste Heat Recovery in the Indian Cement Industry”, Confederation of Indian Industry, Final Report, April 2009 (CII 2009). * MTPA – million metric tons per annum. ANNEX B – BEST PRACTICES IN WHR DESIGN AND OPERATION 43 Variations in preheater exhaust temperature are normally TABLE B-3. TYPICAL AVAILABLE HEAT AND POWER within +/-5oC during normal operation, assuming other GENERATION FROM PREHEATER/GRATE CLINKER COOLER process parameters such as oxygen content and kiln feed remain constant. Clinker cooler temperatures, on the 5000 tpd clinker line, 100% utilization of available other hand, can vary +/-20oC and can exceed 350oC. The waste heat heat availability is also affected by clinker dust carry-over, Input Heat to PH and AQC Boilers 0.963 GJ / Tonne clinker (230 kcal/kg) clinker bed thickness, and variations in clinker quality. Output Heat in Boiler Exhaust Gas 0.379 GJ / Tonne clinker (90 kcal/kg) Note that many plants have two or more kilns at the same Heat Available for Power (Input – 0.583 GJ / Tonne clinker site. Although the heat availability in any single kiln or Output) (140 kcal/kg) cooler may be below an economic size threshold, the total Power Conversion Efficiency 18 – 25% heat available in multiple lines can often be consolidated Potential Power Generation 6 – 9 MW to support a viable WHR project. Source: Adapted from PENTA Engineering, 2013. Raw Material Moisture Content An additional limiting factor to the heat available for of cement and a clinker factor of 0.75, approximately 20 effective recovery is the moisture content of the raw to 30 percent of the required electricity for the cement material entering the kiln. Moisture content of limestone production process can be generated from the waste heat. deposits can range from 2 to 15 percent depending on Figure B-2 shows the band of expected power generation the limestone origin. The amount of moisture present in for a range of kiln capacities. the feed material entering the kiln preheater influences specific thermal energy consumption in the kiln and the Heat recovery and heat transfer ultimately are a function of kiln production rate. Typical practice is to limit moisture the quantity of hot exhaust and the temperature difference content entering the kiln to less than 1.0 percent41. To between the two fluids entering the waste heat boiler. achieve this level, raw feed material is normally dried The improved design characteristics of preheater stages during grinding by utilizing preheater exhaust gas and/or with more effective heat transfer have resulted in lower cooler exhaust as the heat source. exhaust temperature – exhaust temperatures in a modern five-stage preheater can vary between 290 to 320oC, Theoretically about 2.26 GJ is required to evaporate or depending on capacity utilization, operating efficiency remove one tonne of moisture from raw feed or limestone and dust concentration, opening opportunities for organic (540 kcal/kg water). However, in practice, vertical roller Rankine cycle systems capable of producing power from mills require 3.77 to 4.61 GJ of heat per tonne of moisture lower temperatures (as low as 150oC). removed (900 to 1100 kcal/kg water), and ball mills require about 3.14 to 3.56 GJ of heat per tonne of moisture due to losses in mill outlet gas, radiation losses, and air FIGURE B-2. WASTE HEAT POWER GENERATION CAPACITIES AS A FUNCTION OF KILN CAPACITY infiltration (750 to 850 kcal/kg water). To illustrate the impact of moisture on drying requirements, Table B-4 gives the heat requirements in terms of kcal//kg clinker for different limestone moisture levels based on the following assumptions: • Raw meal to clinker factor of 1.55. • Heat requirement of 3.98 GJ / ton of water for raw mill (950 kcal/kg). As shown in Table B-4, substantial heat can be required to dry raw material with high moisture levels. Table B-5 41 “Desk Study on Waste Heat Recovery in the Indian Cement Industry,” Confederation of Indian Industry, Final Report, April 2009. 44 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY shows the heat available in preheater exhaust at different Influence of Dust on WHR preheater exit temperatures ranging from 280°C to 400°C. A comparison of the two tables shows that high moisture Dust in the hot exhaust from the preheater affects both content in the raw feed can significantly reduce the heat kiln efficiency and operation of the waste heat recovery available for WHR in the preheater exhaust; raw feeds with boilers. As shown in Figure B-3, raw material feed typically very high moisture rates essentially eliminate the potential enters the riser duct (connecting duct) coming from the for effective heat recovery. As shown, if the preheater exit Stage 2 cyclone preheater and is carried by the gas to temperature is less than 340°C (heat available of 0.77 GJ/ the top stage cyclone where it is separated. Raw material ton), waste heat is available for recovery only if the moisture dust is carried downward and the gas from Cyclone 1 goes content of the raw feed is less than 10 percent (<0.75 GJ/ to the preheater fan. Depending on the efficiency of the ton heat required for drying). Waste heat is available for top stage cyclone, part of the kiln feed remains in the the entire range of preheater exit temperatures if the raw gas stream. This adds to the specific heat consumption feed has a moisture content of 4 percent (0.28 GJ/ton heat as heated material leaves the kiln system and also required for drying). Raw material moisture content plays a increases the power consumption of the preheater fan major role in the viability of applying WHR by affecting the as the presence of dust increases both the density and heat available from the preheater. Care must be taken to pressure drop. Dust also results in formation of coating in understand the variations in raw material moisture that is the preheater fan impeller during operation due to it clay likely to be encountered over the year, and to base system content. design and performance projections on a reasonable baseline. Cyclone efficiency can vary from 88 percent to as high as 97 percent, resulting in dust concentration varying 0.125 kg/ Another area of drying requirements are cement additives. Nm3 to 0.035 kg/Nm3 43 . The normal range of operation is Wet fly ash and slag specifically often need to be dried 0.070 to 0.080 kg/Nm3. Preheater dust is sticky by nature during grinding, typically with clinker cooler exhaust42: and can affect WHR performance by decreasing the heat transfer rate by forming a coating over the heat transfer • Wet fly ash – 0.40 GJ of heat for every ton of cement surfaces in the waste heat recovery boiler, which in turn based on 30 percent wet fly ash addition and 25 affects the efficiency of the cycle and may eventually percent moisture in fly ash and an evaporation rate cause blockage. To address these conditions, waste heat of 3.98 GJ/ton water. boiler tubes are usually in a vertical arrangement, and • Slag – 0.27 GJ heat for every ton of cement based suitable cleaning systems should be installed to avoid on 50 percent wet fly ash addition and 12 percent dust adherence to the tube walls. Available cleaning moisture in fly ash and an evaporation rate of 3.98 systems are sonic cleaners, soot blowers, mechanical (or GJ/ton water. pneumatic) rapping, and steel shot cleaning (the last two being the most commonly employed). Preheater dust can 42 43 “Desk Study on Waste Heat Recovery in the Indian Cement Industry,” “Desk Study on Waste Heat Recovery in the Indian Cement Industry”, Confederation of Indian Industry, Final Report, April 2009. Confederation of Indian Industry, Final Report, April 2009. TABLE B-4. HEAT REQUIRED FOR RAW MATERIAL DRYING Moisture Content 2% 4% 6% 8% 10% 12% 14% 16% Heat Required, GJ/ton 0.14 0.28 0.43 0.58 0.75 0.92 1.09 1.28 Source: “Manual on Waste Heat Recovery in the Indian Cement Industry”, Confederation of Indian Industry, 2009. TABLE B-5. HEAT AVAILABLE AT DIFFERENT PREHEATER EXIT TEMPERATURES Exit Temperature, °C 260 280 300 320 340 360 380 400 Heat Available, kcal/kg 140.4 151.2 162.0 172.8 183.6 194.4 205.2 216.0 Source: Based on “Manual on Waste Heat Recovery in the Indian Cement Industry”, Confederation of Indian Industry, 2009. ANNEX B – BEST PRACTICES IN WHR DESIGN AND OPERATION 45 As outlined above, the most important information FIGURE B-3. DUST COLLECTION IN PREHEATER/ required to perform an accurate assessment of power CYCLONE TOWERS production potential is the heat available for waste heat recovery from kiln exhaust and cooler air. The remaining information is more relevant in the next stages of basic and detail engineering of the power plant. Therefore, it is necessary to collect a comprehensive set of data records including: • Gas temperature. • Gas composition (and dust content). • Gas flow rates. As noted earlier, the heat available for waste heat recovery is the difference between the heat content in the exhaust of the preheater and the cooler air, minus the heat required for drying or heating raw material and/or fuel also be reduced by installing a high-efficiency cyclone at (coal, heavy fuel oil, refused derived fuel, etc.), as this use the top stage or improving the efficiency of the existing of heat has priority over power production, and provides cyclones. a clear advantage in terms of increased process efficiency (internal heat needs should, in general, be preferred to Cooler exhaust contains dust, which is highly abrasive to power production). steel and capable of eroding the heat transfer area in the boiler, negatively affecting system operation. Waste heat Accurate measurement and analysis of exhaust recovery boilers for cooler exhaust typically use horizontal temperatures, flows, and composition is key to determining tube arrangements and generally do not require cleaning available waste heat. Gas temperatures and flows can systems. A portion of the duct content settles inside the be directly and easily measured on-line, downstream of waste heat boilers, reducing the dust load at downstream all internal heat use for waste heat. Preheater exhaust filters or electrostatic precipitators. Bottom dust removal gas composition can be estimated if an on-line analyser is usually required and typically done by screw conveyors is present, for example, at the stack. If fresh air is mixed or drag chain conveyors. Many installations install a dust upstream of the analyser, the mixing proportion must separation system on the cooler gas before it enters the be known and weighted quantities of components must waste heat boiler. be subtracted from the gas composition at the stack. Clinker cooler air composition is easier to evaluate, as it Project Feasibility Analysis is normally just affected by moisture content. Gas flows All projects should start with precise analysis of all can be calculated indirectly by measuring flows at stack. aspects and factors involved in a potential installation, Again, if fresh air is mixed upstream of the measurement and include technical evaluations of the viability of WHR point, the mixing proportion must be known, and fresh air implementation at the site. The issues and variables to be quantity must be subtracted from the total measure. considered are typically: WHR Project Risks • The size of the plant. There are a number of potential risks related to • Process characteristics and plant layout. implementing WHR that could impact the ability of projects • Heat available from exhaust gas (from kiln and clinker to meet projected performance levels. These risks can be cooler). summarized by the following categories: • Power supply situation (availability of grid power, presence of gen-sets in constant operation or used during power interruption or shortage, etc.). 46 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY TABLE B-6. DESIGN PHASES AND INFORMATION REQUIRED FOR EACH PHASE DESIGN PROCESS PHASES, OBJECTIVES, INFORMATION REQUIRED DESIGN PHASE OBJECTIVES INFORMATION REQUIRED Feasibility assessment • Collect basic process information • Plant size and production capacity and • Evaluate WHR potential operation data • Evaluate power output • Potential reuse of heat (to increase the • Compare technologies and performances production efficiency) • Temperature, flow rate, chemical composition, dust content of exhaust gases available for WHR • Maximum and minimum temperature of exhaust gases acceptable after WHRU • Ambient conditions (localization, height above sea level, climate data) • Potential use for low-grade heat • Plans for plant revamping or other modifications • Market and production forecast Basic design • Define the budget • Plant layout • Define and size the main components for • Tie-in details WHR system • Local regulations to fulfil (for pressure parts, • Draft a preliminary layout environmental, seismic, grid code, firefighting, • Define the requirements for WHR system health and safety) (water, fuel, devices, dedicated personnel) • Company specifications for civil, mechanical, • Draft a procurement and erection programme electrical supply and works Detail design • Finalize the layout of WHR system • Calendar of plant stops for programmed • Finalize the design of WHR system maintenance components • Finalize the procurement and erection programme Design Risk Construction Risk The risks related to system design center on whether the The risks related to construction center on whether the waste heat recovery system is adequately designed and waste heat recovery system is installed according to engineered to achieve the projected performance levels specifications, the overall project schedule is realistically outlined in the project agreement and to withstand the planned and followed, tie-ins to and alterations to existing rigors of extended operation in the industrial environment plant equipment are scheduled around planned downtime, of an operating cement plant. Minimizing these risks and start-up and commissioning are adequately supported. depends on careful vetting of the equipment suppliers and designers and of the local construction content in terms Typical construction times can be as short as six months. of previous experience and any performance guarantees. Installation should be scheduled to ensure there is no It is also critical that the design be based on a complete interruption to cement production. However, individual understanding of the expected operating conditions construction schedules can vary depending on plant including: raw material and fuel composition and moisture conditions and supplier schedules. content; accurate exhaust gas flow, temperature, and composition measurements and/or estimates under The commissioning process can last four to six weeks and varying operating conditions; potential changes in kiln includes check-out of the waste heat recovery system and operation and/or product mix over time; and projections of controls. Commissioning should also include a one- to fuel and electricity prices over time. two- week performance test before final sign-off. System power production should be measured and compared ANNEX B – BEST PRACTICES IN WHR DESIGN AND OPERATION 47 to the design targets (kWh per ton of clinker). While the up to optimum operating conditions after initial installation. performance test is intended to verify that design targets This is a normal operational issue for initial startup of can be met (performance goals sometimes need to be any complicated thermal system, as it takes time for the adjusted to reflect the actual operating conditions of the operating staff to get a feel for the system and its response production line at the time of testing), it should also be to changes in the operating conditions of the kilns. This understood that it normally takes a number of months becomes a risk to achieving projected performance only before the systems are routinely operating at highest if the ramp-up periods are extensive (greater than six efficiency and power output over the range of kiln months). conditions experienced in daily operation. The more critical risk has to do with the month-to-month, Operational Risk or day-to-day, variations in kiln operating conditions. Projected project performance is normally based on design The risks related to waste heat recovery system and kiln heat recovery system power generating capacities at operations center on whether adequate training, operating nominal kiln conditions (nominal exhaust gas temperatures procedures, and maintenance programs are in place to and flows). Variations in either exhaust temperatures ensure that the system performs as expected over the or flow rates impact the amount of heat available for long term. recovery in the waste heat boilers and, in turn, the amount of power produced by the heat recovery systems. Actual Operational risks relate also to the impact on system heat recovery system output could be higher or lower performance by planned or unplanned outages of the than design conditions depending on a variety of factors cement production line or the waste heat recovery system that affect kiln operation. As noted earlier, the factors that itself, and to the ability of the integrated kiln and waste could impact exhaust temperature and flows, and the heat heat recovery system to operate at levels or conditions available for recovery, include: as projected in the PDDs. Along with performance guarantees that are verified during commissioning, most • Fuel quality and moisture content. suppliers provide a system availability guarantee (often • Raw-material quality, temperature, and moisture around 97%). Availability is defined as the ability of the content. system to operate during the time period it is expected • Modifications to kiln operations to maximize to operate (that is, when the cement production line is throughput, enhance fuel utilization efficiency, or operating)44. Similarly, target availabilities for the clinker maintain product quality. production lines also must be considered, including • Changes in operating conditions or inconsistent annual planned maintenance shutdowns and a review of operation related to process control or maintenance historical unplanned outages. issues. Finally, there is the risk of not achieving the full power WHR design is often based on best-case scenarios – high- production levels projected in project design. Performance quality fuel, nominal limestone conditions, set production projections are based on specific design conditions of both mix, etc. However, variations should be expected as the the cement kilns and the waste heat recovery systems. primary objective of the kiln operator is to optimize kiln There are two conditions where the projected production operation for a given fuel composition, clinker quality, and levels or design conditions may not be achieved. One production level. System design should incorporate an reflects the time needed for a system to gradually come understanding of potential off-design conditions likely to 44 Availability Factor (AF) measures on a percent basis the unit’s “could occur, and the potential impact on power production. run” capability. Impacted by scheduled outage hours (SOH) and Forced Outage Hours (FOH). AF = (Scheduled Run Hours - SOH - FOH)*100 / Scheduled Run Hours. 48 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY WHR Maintenance Requirements TABLE B-7. MAINTENANCE REQUIREMENTS FOR DIFFERENT FREQUENCIES Boiler Checklist   Maintenance Frequency Description Comments Daily Weekly Monthly Annually Overall visual inspection Complete overall visual inspection to be sure all equipment is operating and safety x   systems are in place Follow manufacturer’s Compare temperatures with tests performed recommended procedures in after annual cleaning x   lubricating all components Check steam pressure Variation in steam pressure as expected under different loads; wet steam may be x   produced if the pressure drops too fast Check unstable water level Unstable levels can be a sign of contaminates in feedwater, overloading of x   boiler, equipment malfunction Check motor condition Check for proper function temperatures x   Boiler blowdown Verify the bottom, surface, and water column blowdowns are occurring and are effective x   Boiler logs Keep daily logs on: • Exhaust gas temperature • Makeup water volume • Steam pressure, temperature, and amount x   generated Look for variations as a method of fault detection Check oil filter assemblies Check and clean/replace oil filters and x   strainers Check boiler water treatment Confirm water treatment system is functioning properly x   Check flue gas temperatures and Measure flue gas composition and composition temperatures at selected situation x   Check all relief valves Check for leaks x   Check water level control Stop feedwater pump and allow control to stop fuel flow to burner. Do not allow water x   level to drop below recommended level. Inspect system for water/steam Look for: leaks, defective valves and traps, leaks and leakage opportunities corroded piping, condition of insulation x   Inspect all linkages on waste gas Check for proper setting and tightness x   dampers and valves Check all blower belts Check for tightness and minimum slippage. x   Check all gaskets Check gaskets for tight sealing; replace if do x   not provide tight seal Inspect boiler insulation Inspect all boiler insulation and casings for hot spots x   Steam control valves Calibrate steam control valves as specified by manufacturer x   Perform water quality test Check water quality for proper chemical x balance   ANNEX B – BEST PRACTICES IN WHR DESIGN AND OPERATION 49 Boiler Checklist   Maintenance Frequency Description Comments Daily Weekly Monthly Annually Inspect and repair refractories on Use recommended material and procedures x boiler side Relief valve Remove and recondition or replace x Feedwater system Clean and recondition feedwater pumps. Clean condensate receivers and deaeration x system Electrical systems Clean all electrical terminals. Check electronic controls and replace any defective x parts. Hydraulic and pneumatic valves Check operation and repair as necessary       x Turbine Checklist  Maintenance Frequency Description  Daily Weekly Monthly Annually Conduct visual inspection of the unit for leaks (oil and steam), unusual noise/vibration, plugged filters, or abnormal operation x       Cycle non-return valves x   Trend unit performance and health. Hand-held vibration readings should be taken from the steam turbine and gearbox if permanent vibration monitoring x   system is not installed Test emergency backup and auxiliary lube oil pumps for proper operation x   Test the main lube oil tank and oil low-pressure alarms x   Test the simulated over-speed trip if present x   Cycle the main steam stop or throttle valve x   Cycle control valves if steam loads are unchanging x   Cycle extraction/admission valves if steam loads are unchanging. x   Sample and analyze lube oil and hydraulic fluid for water, particulates, and contaminants x   Deferred weekly tests or valve cycling that experience has indicated sufficient reliability to defer them to a one month interval. x   Conduct visual inspection and functional testing of all stop, throttle, control, extraction, and non-return valves including cams, rollers, bearings, rack and pinions, servomotors, and any other pertinent valves or devices for wear, x damage, and/or leakage. Conduct visual inspection of seals, bearings, seal and lubrication systems (oil and hydraulic), and drain system piping and components for wear, leaks, vibration damage, plugged filters, and any other kinds of thermal or mechanical x distress. Conduct visual, mechanical, and electrical inspection of all instrumentation, protection, and control systems. Includes checking alarms, trips, filters, and x backup lubrication and water cooling systems 50 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY Turbine Checklist  Maintenance Frequency Description  Daily Weekly Monthly Annually Test the mechanical over-speed for proper operation annually unless the primary system is electronic and has an OS test switch. For that system, electronic over- speed simulations should be conducted weekly while mechanical and electrical x over-speed tests should be conducted every three years. For electronic systems without an OS test switch, an over-speed test should be conducted annually. Conduct visual inspection of gearbox (if installed) teeth for unusual wear or x damage, and gearbox seals and bearings for damage. Internally inspect non-return valve actuators for wear       x Steam Traps   Maintenance Frequency Description Comments Daily Weekly Monthly Annually Test steam traps weekly/monthly/annually test recommend- ed for low-pressure trap   x x x Repair/replace steam traps When testing shows problems. Typically, traps should be replaced every 3-4 years.       x Cooling Tower   Maintenance Frequency Description Comments Daily Weekly Monthly Annually Overall visual inspection Complete overall visual inspection to be sure all equipment is operating and safety systems x       are in place Inspect for clogging Make sure water is flowing in tower x   Fan motor condition Check the condition of the fan motor through temperature or vibration analysis and compare x   to baseline values Test water samples Test for proper concentrations of dissolved solids, and chemistry. Adjust blowdown and x   chemicals as necessary. Operate make-up water Operate switch manually to ensure proper float switch operation x   Vibration Check for excessive vibration in motors, fans, and pumps x   Check tower structure Check for loose fill, connections, leaks, etc. x   Check belts and pulleys Adjust all belts and pulleys x   Check lubrication Assure that all bearings are lubricated per the manufacture’s recommendation x   Check drift eliminators, louvers, Look for proper positioning and scale build up and fill x   Clean tower Remove all dust, scale, and algae from tower basin, fill, and spray nozzles x Check bearings Inspect bearings and drive belts for wear. Adjust, repair, or replace as necessary.       x ANNEX B – BEST PRACTICES IN WHR DESIGN AND OPERATION 51 Building Automation Systems   Maintenance Frequency Description Comments Daily Weekly Monthly Annually Overall visual inspection Complete overall visual inspection to be sure all equipment is operating and safety systems x       are in place. Verify control schedules Verify in control software that schedules are accurate for season, occupancy, etc. x   Verify set-points Verify in control software that set-points are accurate for season, occupancy, etc. x   Time clocks Reset after every power outage x   Check all gauges Check all gauges to make sure readings are as expected x   Control tubing (pneumatic Check all control tubing for leaks x   system) Check set-points Check set-points and review rational for x   setting Check schedules Check schedules and review rational for x   setting Check dead bands Assure that all dead bands are accurate and that the only simultaneous heating and x   cooling is by design Check sensors Conduct thorough check of all sensors – temperature, pressure, humidity, flow, etc – x   for expected values Calibrate sensors Calibrate all sensors: temperature, pressure, humidity, flow, etc.       x 52 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY ANNEX C – Key Assumptions for the WHR Financial Model and Performance Analysis Financial Analysis of Existing WHR • Auxiliary power requirements were set at an average value of 10 percent of total WHR power generation. Systems This is a conservative estimate, considering a wide An initial modeling exercise estimated the financial variation in values observed in the survey, with the performance of the existing WHR systems included in the lowest value reported of 5.5 percent of total power survey under two financial analysis cases: generation. • Information on annual O&M costs was not provided 1. Planned Performance – operations based on design by all companies. For the purpose of consistency, an conditions, including the maximum planned operating estimate was developed based on partial reporting hours of clinker lines in a given year with 100% WHR and industry best practices: annual O&M costs equal capacity utilization. to $70/kW for systems <7.5 MW, $60 /kW for systems 2. Actual Performance – operations based on reported >7.5 MW and <10 MW, and $50 /kW for systems >10 actual operating hours of clinker lines and actual MW. WHR capacity utilization in a given year. • Performance of WHR systems during the first year of operations was set to 94 percent in the model, Key assumptions used in the financial analysis include: based on industry experience that WHR plants may be operating at 85-92 percent of capacity during the • The data used were collected in a survey, with first three to six months until the system operations questionnaires completed by individual companies are optimized. based on their operations in 2015 to ensure • For the financial analysis, equity-only financing consistency within the group. No adjustments were was assumed for all companies to ensure ease of made to primary data received. In case no data comparisons. Consequently, a discount rate of 12.5 was provided for select questions, estimates were percent was applied to all companies. It was further made based on other relevant plant parameters assumed that the full investment cost was paid within and responses of comparable companies within the one year prior to start-up. sample. • Economic lifetime for WHR systems was set to 20 • Planned clinker kiln operating hours were based on years in the model. Best practices in operations the maximum annual operating hours reported for and maintenance can extend the lifetime of WHR each plant. Actual kiln operating hours were based systems up to 30 years. Yearly degradation rate is on the reported kiln operating hours for 2015. highly dependent on quality of operation and was not • Planned WHR operating hours were estimated considered in the analysis. using a 100 percent WHR system availability, as • The cost of electricity paid by cement plants in each a percent of maximum kiln operating hours, as given year was assumed at the level of yearly average typically guaranteed by the suppliers. In a plant wholesale market prices in Turkey during that year, operating 8,000 hours a year this corresponds to including fees for transmission, distribution, and 240 hours of WHR downtime, which is typical given renewable energy support. The rate of increase for the annual hours of planned and unplanned outages the out-years was based on forecasts in Mercados for maintenance and service. Actual WHR operating Market Report of March 2017. hours were based on the reported WHR system operating hours for 2015. ANNEX C – KEY ASSUMPTIONS FOR THE WHR FINANCIAL MODEL AND PERFORMANCE ANALYSIS 53 • Historical and forecasted values of exchange rates Key assumptions used in the financial analysis include: (Turkish lira vs the US dollar and euro) and inflation were based on Economist Intelligence Unit estimates, • The data used were collected in a survey, with accessed on July 21, 201745. questionnaires filled by individual companies based on their operations in 2015 to ensure consistency Sensitivity Analysis of ORC WHR for within the group. No adjustments were made to primary data received. In case no data was provided Plants without Existing WHR Systems for select questions, estimates were made based A sensitivity analysis was conducted to identify the key on other related plant parameters and responses of performance and financial factors that could impact the comparable companies within the sample. financial performance of potential ORC WHR projects in • The potential for additional WHR application was the Turkish cement industry. In the analysis, a Base Case calculated assuming the start-up of new WHR plants technical feasibility and financial analysis was conducted in early 2018. for each plant based on their reported operating data for • Base Case clinker kiln operating hours were based 2015, including information on kiln utilization and operation, on the maximum annual kiln operating hours reported exhaust temperatures and flows from preheaters and for each plant. clinker coolers, and the amount of heat used for drying • Base Case WHR operating hours were based on of raw material and fuel. ORC WHR systems were sized a WHR system availability of 96 percent applied for each plant based on the preheater and clinker cooler to maximum kiln operating hours. In a plant exhaust conditions, and technical and financial feasibility operating 8000 hours a year, 96 percent availability was estimated based on typical cost and performance corresponds to 320 hours of WHR system downtime, assumptions for ORC systems applied to cement plants. which is typical of the annual hours of planned and A number of scenarios were then conducted to assess the unplanned outages for maintenance and service. financial sensitivity of ORC projects to variations in key Suppliers of ORC suppliers generally guarantee and plant and ORC system operating and financial parameters availability of 95-98 percent. as described in Table C-1. • Heat available for waste heat recovery was calculated based on average temperatures and flowrates of the 45 Economist Intelligence Unit, http://www.eiu.com/home.aspx. exhaust gases and air at the pre-heater and clinker cooler exits as reported by the companies surveyed. • The composition of gases assumed was in line with industry practice at the preheater with 64% TABLE C-1. DESCRIPTION OF FINANCIAL ANALYSIS N2, 28% CO2, 4% H20, 4% O2 and clinker cooler SCENARIOS FOR ORC WHR with 72% N2 and 21% O2. The normal exhaust gas density for preheater exhaust was 1.438 and 1.276 Scenario Description Nm3/kg for clinker cooler exhaust air. Specific heat Base Case Operations as per maximum planned operating hours of cement plant in given year capacities were calculated based on gas density and with 100% WHR capacity utilization, with heat temperature. used for drying limiting available heat for WHR • The temperatures of exhaust gas and air at the as per typical drying practice and as reported by each plant. exit of the WHR heat exchangers applied to the 1 WHR Capacity Utilization: -30%/+5% preheater and clinker cooler were based on usual cement industry practice. A temperature of 230°C 2 Clinker Kiln Operating Hours: 6034 hours, 8595 hours at the exit of preheater and/or cooler was assumed 3 Total Capital Cost: -10% / +20% where heat was being used for raw material or fuel 4 Electricity Prices: +10% / -10% drying as reported in the survey. In case of limited 5 TL/$ Exchange Rate: -10% / +10% or no secondary use of heat after the WHR, 160°C temperature was assumed at the exit of the preheater 6 WHR O&M Expenses: -130% / +30% recovery unit (the minimum temperature required to 7 WHR Auxiliary Power: 8%, 10% (base case), 12% stay above condensation temperature for combustion gases to reduce heat exchanger corrosion); 120°C 54 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY was assumed at the exit of the cooler recovery • Cost of electricity paid by cement plants in each unit since the exhaust is primarily air and does not given year was assumed at the level of yearly contain condensable corrosive elements. The ability average wholesale market prices in Turkey during to recover heat from streams as low as 120°C is one that year, including fees for transmission, distribution, of the advantages of ORC systems. and renewable energy support. The rate of increase • Heat transfer losses were assumed at 2%, in line with for the out-years is based on forecasts in Mercados industry practice. Gross electric efficiency of the ORC Market Report, March 2017. unit was set to 21%, which is an upper-range value for • Historical and forecasted values of exchange rates ORC plants currently in place, considering that ORC (Turkish lira vs. U.S. dollar and euro) and inflation solutions of higher efficiency are under development were based on Economist Intelligence Unit estimates, and should be more readily available in the near accessed on July 21, 2017. future. • Auxiliary power requirements tend to vary between 9 Levelized Cost of Electricity and 12 percent based on site and design conditions; Levelized cost of electricity (LCOE) is a convenient 10 percent was assumed for the Base Case. measure of the overall competitiveness of a generation • Total investment costs were estimated based on investment compared to other electricity supply options. average market prices in Europe during the first LCOE represents the per-kWh cost (in discounted real quarter of 2017 with respect to system capacity: dollars and discounted kWhs) of installing and operating €3500 per kW for installations of 2-3 MW capacity, a generating asset over an assumed financial life and duty €2915 per kW for 3-5 MW, €2335 per kW for 5-10 cycle or capacity utilization: MW, and €2105 per kW for installations above 10 MW capacity. lifetime (a) initial investment cost (€) Annual costs ($/a) • Annual O&M costs were assumed at 3 percent of (expected- estimated) investment cost for systems of 2-3 MW capacity, 2.5 discount percent for systems 3-5 MW capacity, and 2.0 percent Et rate (%) for systems above 5 MW. These are typical of usual Levelised Cost of output (MWh/a), annual Electricity electricity generation O&M costs in existing applications, and include spare ($/MWh) (expected-estimated) parts, personnel, and the cost of a major overhaul Key inputs to calculating LCOE for a project include capital that can be expected around the mid-life of the plant. costs, fuel costs, operating and maintenance (O&M) • Performance of the WHR system during the first year costs, financing costs, and an assumed utilization rate. of operations was assumed to be 94% in the model, The importance of each of these factors varies among as during first three to six months after the start-up generating technologies. For technologies such as solar WHR plants may be operating at 85-92% of capacity, or WHR generation that have no fuel costs and relatively until the system operations are optimized. low O&M costs, LCOE changes in rough proportion to • Equity-only financing was assumed for all of the the capital cost and utilization rate of the system. For companies, although loans and several incentive technologies with significant fuel cost, both fuel cost programs are currently available to finance WHR and capital cost estimates significantly affect LCOE. The projects in the cement sector in Turkey. A discount availability of various incentives, including regional or rate of 12.5% was applied to all companies, and it national tax credits, can also impact the calculation of was further assumed that the full investment cost was LCOE. As with any projection, there is uncertainty about paid within one year prior to start-up. these factors and their values can vary regionally and • Economic lifetime for WHR systems was set to 20 across time as technologies evolve and fuel prices change. years in the model. Best practices in operations While LCOE can be an important measure in considering and maintenance can extend the lifetime of WHR the economic viability of a WHR project, it is important systems up to 30 years. Yearly degradation rate is to base the calculation on adequate upfront preparation highly dependent on quality of operation and was not and realistic projections of key performance and cost considered in the analysis. parameters . ANNEX C – KEY ASSUMPTIONS FOR THE WHR FINANCIAL MODEL AND PERFORMANCE ANALYSIS 55 56 Detailed Sensitivity Analysis Results TABLE C-2. SENSITIVITY ANALYSIS RESULTS WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY ANNEX D – Subsidies and Financing Mechanisms for WHR Investments Various financing mechanisms are provided by the the local economy, while the amount of support varies international and national institutions and organizations depending on the level of development in the regions. The to support new investments in Turkey. Because there are minimum fixed investment amount is defined separately many incentives for special investments in the industry for each sector and region with the lowest amount being such as WHR, it is useful to review available incentives and TL 1 million in Region 1 and 2, and TL 500,000 in the financial grants before the investment stage. remaining regions. Many of the Incentives vary according to investment Incentives region of the county: First, the company must apply to the Ministry of Economy to get an investment incentive certificate. After that The specific support instruments provided within the application, the ministry prepares an investment incentive framework of the four investment incentives schemes are certificate to support the company for the related summarized in Table C-1; WHR investments would qualify investment(s). The certificate initiates the support for under Regional Investment Incentives Schemes. investors; the level of support varies from region to region. Regional Investment Incentives Scheme VAT Exemption The sectors to be supported in each region are determined The generally applied VAT rate is set at 1 percent, 8 percent, in accordance with regional potential and the scale of and 18 percent. Commercial, industrial, agricultural, and FIGURE D-1. TURKISH INVESTMENT REGION MAP Source: investinturkey.com ANNEX D – SUBSIDIES AND FINANCING MECHANISMS FOR WHR INVESTMENTS 57 independent professional goods and services, goods Interest Rate Support and services imported into the country, and deliveries of Interest support is provided for loans with at least a one- goods and services as a result of other activities are all year term obtained within the frame of the Investment subject to VAT. Incentive Certificate. The interest/dividend can be covered by the government for the portion of the loan that Customs Duty Exemption covers 70% of the investment budget. On 9 May 2014 the Customs duty is not paid for machinery and equipment 2014/6058 Decision in the 28995 Number Official Gazette provided from abroad (imported) within the scope of the further clarified that energy efficiency and waste heat investment incentive certificate. recovery investments are added to the top priority list, and all priority investments will be supported with 5th region Social Security Premium Support (Employer’s Share) incentives (independent of location) under “Incentives and State Grants in Investments Scheme” (see Chapter 3 This measure stipulates that, for additional employment Table-X for detailed analysis). created by the investment, the employer’s share of the social security premium on portions of labor wages References corresponding to the legal minimum wage will be covered by the ministry. 1. Invest in Turkey, http://www.invest.gov.tr/en-US/ investmentguide/investorsguide/Pages/Incentives. aspx. TABLE D-1. SUPPORT INSTRUMENTS WITHIN INVESTMENT INCENTIVES SCHEMES General Regional Large-Scale Strategic Investment Investment Investment Investment Support Instruments Incentives Incentives Incentives Incentives Scheme Scheme Scheme Scheme VAT Exemption X X X X Customs Duty Exemption X X X X X Tax Reduction X X X Social Security Premium Support X X (Employer’s Share) Income Tax Withholding Allowance * X X X X Social Security Premium Support X X (Employee’s Share) * Interest Rate Support ** X X Land Allocation X X X X VAT Refund*** *Provided that the investment is made in Region 6. **Provided that the investment is made in Regions 3, 4, 5, or 6 within the framework of the Regional Investment Incentives Scheme. ***For construction expenditures of strategic investments with a minimum fixed investment amount of TL 500 million. 58 WASTE HEAT RECOVERY IN TURKISH CEMENT INDUSTRY 59