Biomass Resource Mapping in Vietnam PHASE 2 IMPLEMENTATION PLAN DECEMBER 2017 This report was prepared by Full Advantage, Simosol, Institute of Energy and Enerteam, under contract to The World Bank. It is one of several outputs from the biomass Energy Resource Mapping and Geospatial Planning Vietnam [Project ID: P145513]. This activity is funded and supported by the Energy Sector Management Assistance Program (ESMAP), a multi-donor trust fund administered by The World Bank, under a global initiative on Renewable Energy Resource Mapping. Further details on the initiative can be obtained from the ESMAP website. This document is an interim output from the above-mentioned project. Users are strongly advised to exercise caution when utilizing the information and data contained, as this has not been subject to full peer review. The final, validated, peer reviewed output from this project will be the Vietnam Biomass Atlas, which will be published once the project is completed. Copyright © 2017 THE WORLD BANK Washington DC 20433 Telephone: +1-202-473-1000 Internet: www.worldbank.org The World Bank, comprising the International Bank for Reconstruction and Development (IBRD) and the International Development Association (IDA), is the commissioning agent and copyright holder for this publication. However, this work is a product of the consultants listed, and not of World Bank staff. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work and accept no responsibility for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for non-commercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: +1-202-522-2625; e-mail: pubrights@worldbank.org. Furthermore, the ESMAP Program Manager would appreciate receiving a copy of the publication that uses this publication for its source sent in care of the address above, or to esmap@worldbank.org. BIOMASS RESOURCE MAPPING IN VIETNAM REPORT ON PHASE 2 IMPLEMENTATION AND REVISED WORK PLAN FOR PHASE 3 Prepared by: Full Advantage Co., Ltd. (FA), Thailand (Lead Consultant) Simosol Oy and partners from Finland Institute of Energy, Vietnam Energy Conservation Research and Development Center, Vietnam Date: 11 December 2017 Page | 1 Country: Vietnam Project title and ID: Renewable Energy Resource Mapping: Biomass [Phases 1-3] - Vietnam Project ID: P145513 Implementing agency: The World Bank (Vietnam) in close coordination with the General Department of Energy (GDE) under the Ministry of Industry and Trade (MOIT) of Vietnam The Consultant Consortium: Full Advantage Co., Ltd (FA), Thailand (Lead Consultant) Simosol Oy, Finland VTT Technical Research Center of Finland MHG Systems, Finland Wiltrain Oy, Finland PITCO Pvt., Ltd., Pakistan Institute of Energy (IE), Vietnam Energy Conservation Research and Development Center (Enerteam), Vietnam Date of report: 11 December 2017 Page | 2 TABLE OF CONTENTS 1. Project Introduction and Background ............................................................................ 6 2. Expected Outputs and Summary of Achievements of Phase 2 .................................... 7 2.1 Summary of Achievements vs Expected Outputs ............................................................................... 7 2.2 Major Deviations and Mitigations ........................................................................................................... 9 3. Remote Data Collection and Analysis ............................................................................ 9 3.1 Gathering of Satellite Images .................................................................................................................... 9 3.2 Analysis of Satellite Images ..................................................................................................................... 10 4. Training on Data Collection for Enumerators ............................................................. 14 4.1 Development of the Crop Survey Form for MHG Mobile App .................................................... 14 4.2 Development of Survey Questionnaires for Industrial Biomass Residues .................................. 14 4.3 Training Workshop on Field Survey and Data Collection .............................................................. 14 5. Data Collection and Validation ..................................................................................... 15 5.1 Collection and Validation of Crop Biomass Data ............................................................................. 15 5.2 Collection and Validation of Industrial Biomass Data ...................................................................... 21 6. GIS Data of other Driving Components ....................................................................... 24 7. Data Analysis and Development of Draft Biomass Atlas ............................................ 24 7.1 Potential Maps of Crop Harvesting Residues and Power Plant Site Suitability .......................... 24 7.3 Biomass-Based Electricity Production Potential at the Existing Biomass Producing Sites ...... 33 8. Data Validation Workshop ............................................................................................ 42 9. Revised Work Plan for Phase 3 ..................................................................................... 43 10. Conclusions and Recommendations ............................................................................. 46 10.1 Conclusions.............................................................................................................................................. 46 10.2 Recommendations .................................................................................................................................. 47 11. Attachments .................................................................................................................... 47 Page | 3 LIST OF TABLES Table 1: Summary of achievements vs expected outputs of Phase 2 ............................................................. 8 Table 2: Summary of number of districts and farmers surveyed .................................................................. 17 Table 3: Summary of the accepted datasets by industrial sector .................................................................. 23 Table 4: The residue-to-crop ratio (RCR) used in the analysis ..................................................................... 26 Table 5: Summarized results of analysis of 16 sugar mills ............................................................................... 35 Table 6: Summarized results of analysis of 46 rice mills ................................................................................. 38 Table 7: Summarized results of analysis of 10 MSW landfills ......................................................................... 40 Table 8: Summarized results of analysis of 19 livestock farms ...................................................................... 41 Table 9: Revised Work Plan of Phase 3 .............................................................................................................. 44 LIST OF FIGURES Figure 1: Sentinel-1 image tiles used in the analysis. ........................................................................................ 10 Figure 2: Single radar signal backscattering band for the MDR agro-ecological zone in Southern Vietnam ....................................................................................................................................................................... 11 Figure 3: False color image mosaic using backscattering from Nov 2015, Mar 2016 and Aug 2016 for red, green and blue values. ..................................................................................................................................... 12 Figure 4: Sample of the land use classification result in the Mekong delta. ................................................ 12 Figure 5: Sample of the land use classification result in the Mekong delta. ................................................ 13 Figure 6: A reference field sample that was included into land use classification reference sample data set. ................................................................................................................................................................................ 16 Figure 7: An example of a rejected reference field sample due to having been recorded in the middle of a road leaving uncertainty for the actual location of the field. ................................................................. 17 Figure 8: Locations of farms surveyed ................................................................................................................. 19 Figure 9: Locations of farms with collected datasets accepted ..................................................................... 20 Figure 10: Map of the surveyed industrial sites ................................................................................................. 23 Figure 11: Components of the first atlas and harvest residue feedstock available at farm level ........... 25 Figure 12: Draft total technical harvest residue feedstock estimate map ................................................... 27 Figure 13: Steps to create the industrial scale power generation potential atlas...................................... 28 Figure 14: The modeling principle for the different site suitability factors. ................................................ 29 Figure 15: Road and watercourse network data used in the analysis .......................................................... 30 Figure 16: Grid stations, and the computed grid station distance index. .................................................... 31 Figure 17: Site-suitability index for 15 MW circulating fluidized bed combustion based power plant.32 Figure 18: Map of potential high-pressure cogeneration plants at the existing sugar mills..................... 36 Figure 19: Map of potential power plants at 46 existing rice mills in the South of Vietnam .................. 38 Figure 20: Map of potential power plants at the surveyed MSW landfills in the South of Vietnam ..... 40 Figure 21: Map of potential power plants at the surveyed livestock farms in the South of Vietnam ... 42 Page | 4 LIST OF ACRONYMS AHAV Animal Husbandry Association of Vietnam DARD Department of Agriculture and Rural Development (of the provinces) DOIT Department of Industry and Trade (of provinces) DONRE Department of Natural Resources and Environment (of provinces) ESA European Space Agency ESMAP Energy Sector Management Assistance Program FA Full Advantage Co., Ltd. (Thailand) FAO Food and Agriculture Organization GDE General Department of Energy (under MOIT) GIZ Gesellschaft fur Internationale Zusammenarbeit (Germany) GIS Geographic Information System GOV Government of Vietnam IE Institute of Energy (Vietnam) M&E Monitoring and Evaluation MOIT Ministry of Industry and Trade MONRE Ministry of Natural Resources and Environment MSW Municipal Solid Waste NFIS National Forest Inventory and Statistics NLU Nong Lam University NPDP National Power Development Plan PCF Power Capacity Factor PITCO PITCO Private Limited (Pakistan) RE Renewable Energy REDP Renewable Energy Development Project RERM Renewable Energy Resource Mapping SAR Synthetic Aperture Radar S-1 Sentinel-1 TOR Terms of Reference VDA Vietnam Diary Association VNUA Vietnam National University of Agriculture VSSA Vietnam Sugar and Sugarcane Association WB World Bank Page | 5 1. PROJECT INTRODUCTION AND BACKGROUND According to the "National Power Development Plan (NPDP) for the period of 2011-2020 with an outlook to 2030" (referred to as PDP VII)1, the Government of Vietnam (GOV) has set a national target for increasing the total amount of power generation and import from about 19,500 MW in 2010 to 75,000 MW and 146,800 MW by 2020 and 2030, respectively. The total electricity generation and import is expected to be 330 billion kWh in 2020 and 695 billion kWh in 2030. The amount of electricity generated from renewable energy (RE) sources would be around 42 billion kWh in 2030, accounting for 6% of the total amount of electricity generation and import. The revised NPDP (PDP VII-revised) promulgated by the Prime Minister of Vietnam in 20162 has reduced the total electricity demand projection for 2030 from 695 to 572 TWh/year. However, the target for electricity generation from RE sources (excluding large-scale and pumped-storage hydropower plants) was increased from 42 TWh/year to around 61 TWh/year, making up 10.7% of the total electricity generated and imported in 2030. Solar power will account for 3.3% of the total electricity generation and import, followed by small hydropower (3.2%), wind power (2.1%), and biomass and MSW (2.1%). In order to attain such ambitious targets, the GOV has endeavored to exploit various sources of power generation and supply: fossil fuels (coal and gas), hydropower, nuclear power, RE and imported power. As Vietnam has a huge potential of RE resources, the GOV has set a goal to increase the total installed power capacity of RE sources from around 2,400 MW in 2015 to 23,350 MW in 2030. The installed power capacity is expected to be 6,000 MW for wind power, 12,000 MW for solar power, 3,350 MW for small hydro power (with a capacity below 30 MW), and 2,000 MW for biomass (including MSW) by 2030.3 The Ministry of Industry and Trade (MOIT) is implementing the Renewable Energy Development Project (REDP) funded by the World Bank. The objective of the REDP is to increase the supply of electricity to the national grid from RE sources on a commercially, environmentally and socially sustainable basis. The REDP has three components: (i) Investment Implementation; (ii) Regulatory Development and (iii) Project Pipeline Development. MOIT is implementing several technical assistance activities to strengthen the capacity of government agencies and stakeholders for exploiting the sizable RE resources of Vietnam. In addition to studies on supporting mechanisms for development of RE and cumulative impact assessment for cascade hydropower projects, MOIT has requested the assistance of the World Bank for a Renewable Energy Resource Mapping (RERM) project, with funding from the Energy Sector Management Assistance Program (ESMAP), a global knowledge and technical assistance program administered by the WB and supported by eleven bilateral donors. The development objective of this project is to increase the output and diversity of renewable electricity generation in Vietnam. The outcome objective is to improve the awareness of the government and the private sector of the resource potential for biomass, small hydropower, and wind, and providing the government with a spatial planning framework to guide commercial investments. 1 Decision 37/2011/QD-TTg dated 14 June 2011 of the Prime Minister of Vietnam 2 Decision 428/QD-TTg dated 18 March 2016 of the Prime Minister of Vietnam 3 Vietnam: Energy sector assessment, strategy, and road map. Asian Development Bank, December 2015. Page | 6 Under this RERM project, the World Bank has contracted a Consortium of consultants led by Full Advantage Co., Ltd. (FA) to develop a Biomass Atlas for Vietnam (hereafter called “FA Consortium”). The FA Consortium involves several Finnish companies led by Simosol Oy, and two local partners: the Institute of Energy (IE) and the Energy Conservation Research and Development Center (Enerteam). The Vietnam National University of Agriculture (VNUA) and the Nong Lam University (NLU) were contracted by MOIT to conduct the field survey and data collection on crop and industrial biomass residues. The overall objective of this biomass resource mapping project is to support the sustainable expansion of electricity generation from biomass by providing the national government, provincial authorities and commercial developers in Vietnam with an improved understanding of the location and potential of biomass resources. The specific objective is to support RE mapping and geospatial planning for biomass resources in Vietnam. The project consists of three phases:  Phase 1: Project inception, team building, data source identification, preparation of terms of reference (TOR) for field survey and data collection (conducted by the local contractors hired by MOIT), and implementation planning;  Phase 2: Data collection/analysis and creation of draft Biomass Atlas;  Phase 3: Production and publication of a validated Biomass Atlas. The present document is the Progress Report on “Phase 2 Implementation of the Biomass Resource Mapping for Vietnam”. It covers the activities carried out in the period from August 1, 2016 to November 30, 2017. The report also includes the revised Work Plan for Phase 3 of the project. The main activities of Phase 2 consisted in: 1. Remote data collection and analysis; 2. Training on data collection for enumerators; 3. Validation of data collected by enumerators; 4. Acquisition of GIS data of other driving components; 5. Data analysis and development of a draft biomass atlas; 6. Conduct of stakeholder data validation workshop. The implementation of these activities and their outputs are reported in sections 3, 4, 5, 6, 7, and 8 respectively. 2. EXPECTED OUTPUTS AND SUMMARY OF ACHIEVEMENTS OF PHASE 2 2.1 Summary of Achievements vs Expected Outputs The expected outputs and the summary of achievements of Phase 2 of the project are presented in Table 1. Page | 7 Table 1: Summary of achievements vs expected outputs of Phase 2 Activity Expected outputs Achievements 1. Remote data  Remote data collected and analyzed  Satellite images were acquired from Sentinel-1 and were analyzed to produce the raw collection and analysis biomass cluster images for field observation and inspection.  A field inventory plan was created. 2. Training on data  Training on field survey and data  MHG Biomass Manager was developed. collection for collection conducted  Required smartphone applications for navigation, data entry and data transfers were enumerators acquired.  Training on field survey and data collection was conducted on 28-29 September 2016 for 32 participants from Nong Lam University (NLU) and Vietnam National University of Agriculture (VNUA). 3. Data collection and  Field surveys conducted, and data  Field surveys were conducted, and the data on crop biomass residues were collected in ALL validation collected (by the local consultants hired 63 cities/provinces of Vietnam with 21,212 farmers interviewed. Collected data on crop by MOIT) biomass residues were validated, and 19,985 datasets were accepted.  Collected data validated by FA  Field surveys on industrial biomass residues were conducted in 16 cities/provinces in the Consortium South of Vietnam and in 2 other provinces in the central region. 137 datasets from six industrial sectors (sugar mills, rice mills, MSW landfills, livestock farms, wood processing mills and brick-making factories) were collected and validated. 4. Acquisition of GIS  GIS data of other driving components  Power grid sub-station locations (digitized at the World Bank from a paper map), data of other driving (road network, power T&D network, OpenStreetMap road and waterway network data were acquired. components etc.) acquired and verified. 5. Data analysis and  A comprehensive database necessary  The collected data were processed and integrated into a comprehensive database. development of draft for biomass resource mapping,  Draft biomass resource maps were produced. biomass atlas including raw data files elaborated  Draft biomass resource maps developed 6. Conduct of  A stakeholder data validation workshop  A stakeholder data validation workshop was conducted on 15 November 2017 stakeholder data conducted validation workshop Page | 8 2.2 Major Deviations and Mitigations There have been some deviations from the Phase 2 Implementation Plan that was approved by the World Bank (WB) in October 2015. They are as follows:  Delay in hiring local consultants for field survey and data collection: According to the approved Phase 2 Implementation Plan, Phase 2 should have started in November 2015. However, due to a longer time needed by MOIT to hire the local consultants for the field survey and data collection, the kick-off of Phase 2 implementation was delayed for almost a year. Phase 2 could actually start in September 2016 with a training workshop on “Field Survey and Data Collection” for the two selected local consultants (NLU and VNUA).  Delay in field data collection: The industrial survey in the northern and central provinces has been delayed due to several reasons, the main one being difficulties to identify a proper source of funding for that activity. 3. REMOTE DATA COLLECTION AND ANALYSIS 3.1 Gathering of Satellite Images The first step in the remote data collection and analysis phase was gathering the satellite images. In order to achieve the goals, it was decided that 13 epochs (images from different times of the year) encompassing the entire territory of Vietnam would be needed. Each epoch targeted the middle of the month from November 2015 to November 2016. In total, the analysis used 617 images. Figure 1 shows the coverage of the images superimposed over the map of Vietnam. Due to prevailing cloudiness over the country, synthetic aperture radar (SAR) images from the Sentinel-1 (S-1) satellite by the European Space Agency (ESA) were used. The benefit of SAR images is that the radar signal is an active signal sent by the satellite which penetrates the cloud cover to reach the ground. This increases the availability of images over any given period considerably compared to the optical satellite images, especially during the monsoon season. The S-1 images were downloaded from the Copernicus Open Access Hub4. 4 https://scihub.copernicus.eu/ Page | 9 Figure 1: Sentinel-1 image tiles used in the analysis. 3.2 Analysis of Satellite Images The satellite image processing consisted of 5 stages described below. 1. Image preprocessing: the preprocessing stage extracts the satellite image data and its metadata from the distribution archive downloaded from Copernicus Open Access Hub. Additional stages in preprocessing are radar signal backscattering noise reduction, radiometric calibration, and terrain correction. This was executed with the Sentinel-1 Toolbox5 pre-processing software developed and provided by ESA. 2. Image mosaicking: After the preprocessing steps, the images having 20x20 m ground resolution were mosaicked, i.e. "stitched" together into one seamless image covering the whole analysis area. This is illustrated in Figure 2, showing one backscattering band of a single monthly image mosaic. 5 http://step.esa.int/main/toolboxes/sentinel-1-toolbox/ Page | 10 Figure 2: Single radar signal backscattering band for the MDR agro-ecological zone in Southern Vietnam 3. Image stacking: The two-band monthly image mosaics were then "stacked on top of each other" to form the 26-band analysis image (2 bands per image times 13 monthly images). Figure 3 shows a false color rendition of the stacked images using the first three of the 26 bands for the red, blue and green bands of a regular RGB image. 4. Ground reference data processing: individual field observations recorded with the MHG Mobile software during the field survey (see below) were processed into “per cropping season reference observations” (geographic coordinates for the observation, cultivated crop, or fallow period, for each of the three cropping cycles during a year). 5. Land use classification: We trained a land use Random Forest classifier model based on the multi-temporal S-1 image stack and the reference observations. This classification model was then used to predict the land use class for all the pixels in the S-1 image stack. The modeling and prediction was done separately for each of the nine agro-ecological zones6 of Vietnam. The crops for the land use classification were: rice, maize, peanut, cassava, coffee, sugarcane, tea, pepper, rubber tree, cashew nut, mango, rambutan, longan, orange, coconut, mandarin and other (all other crops pooled as a single class). The other land use classes were fallow, waterlogged field, bare area, water, urban, bush and forest. Figure 4 shows a sample of the classification result. 6 http://www.fao.org/ag/agp/agpc/doc/counprof/vietnam/vietnam.htm Page | 11 Figure 3: False color image mosaic using backscattering from Nov 2015, Mar 2016 and Aug 2016 for red, green and blue values. Figure 4: Sample of the land use classification result in the Mekong delta. Page | 12 Figure 5 shows the classification accuracy for the MRD agro-ecological zone in the Mekong Delta. Overall, the classification accuracy is good, and the non-agricultural areas are cleanly separated from the agricultural areas. However, three problems are also clearly visible; (i) overrepresentation of the most common land use classes in the expense of the less prominent classes (other classes classified as rice, the most prominent class), (ii) misclassification between permanent, woody crops (i.e., coconut, rambutan, mango, etc.), and (iii) misclassification between the "other" crop and rest of the as well as urban and forest classes. Figure 5: Sample of the land use classification result in the Mekong delta. Therefore, the remaining steps for the finalization of the biomass atlas during Phase 3 are:  Testing of additional optical satellite images to improve crop distinction,  Addition of existing land use GIS data to delineate the analysis area better,  Modelling improvements to reduce the over-estimation of dominant crop classes for each agro-ecological zone. Page | 13 4. TRAINING ON DATA COLLECTION FOR ENUMERATORS 4.1 Development of the Crop Survey Form for MHG Mobile App A smartphone based field survey questionnaire using the MHG Mobile app 7 was developed and implemented. The form had sections for questions about the respondent, farm, crops cultivated on the farm, harvest residue marketing efforts, and finally about a particular reference field for the satellite image analysis. The detailed composition of the survey form is documented in Attachment 1. To accompany the mobile field survey form, a data validation routine was developed where the survey data was daily downloaded from the backend system of MHG Mobile, stored in an Excel file containing all the survey data collected for that day. A validation routine was also run on the data being written to the Excel file. In case any invalid data was detected, the Excel file had the invalid data highlighted. The reasons for non-validity was written in the file as well. The same validation rules, mainly relating to crop yield levels, and harvest residue utilization pattern data, were also applied a second time when the district level final survey results were compiled. The validation rules were fine-tuned over the progress of the survey. 4.2 Development of Survey Questionnaires for Industrial Biomass Residues FA Consortium developed the survey questionnaires for seven industrial sectors:  Sugar mills,  Rice mills,  MSW landfills,  Livestock farms,  Wood processing mills,  Brick-making factories,  Pulp and paper mills. These survey questionnaires are provided in Attachment 2. 4.3 Training Workshop on Field Survey and Data Collection A training workshop on field survey and data collection was conducted at NLU in Ho Chi Minh City on 28-29 September 2016. The objective of the training workshop was to provide hands-on training to the field enumerators of the local consultants (NLU and VNUA) for conducting surveys using 3G smartphones and Android applications designed and recommended by Simosol, Finland. The workshop program was divided into two sessions spread over two days. On the first day, a classroom training session was conducted by FA Consortium. On the second day, a pre-testing field activity in the morning was organized to test the survey methodology. A question and answer session with an assessment of the exercise was conducted in the afternoon. 7 https://play.google.com/store/apps/details?id=com.mhgsystems.ui Page | 14 Thirty-two participants (10 from VNUA and 22 from NLU) were trained. A pre-testing field activity was organized in Cai Lay town in the Tien Giang province. The Department of Natural Resources and Environment (DONRE) of Tien Giang province was contacted by NLU to coordinate with the management of two rice mills for the industrial biomass survey, and to contact five farmers within the vicinity of the town for the crop biomass survey. The completion report on this training workshop is provided in Attachment 3. 5. DATA COLLECTION AND VALIDATION 5.1 Collection and Validation of Crop Biomass Data 5.1.1 Methodology of Crop Biomass Survey As agreed with WB and MOIT during the Inception Mission, while the field survey sampling will cover the whole country, more samples will be collected in areas with abundant biomass. The resulting Biomass Atlas will provide more details on those areas and fewer details on areas with scarce biomass resource. No field survey was conducted for forested areas. Instead the dataset that was planned to be used for the baseline biomass information is the output of the National Forest Inventory (NFIS) project executed by the forest authority VNFOREST in 2014-2016. Currently, an open data sharing policy is being implemented at the FORMIS II project in VNFOREST. Simosol has already accessed a sample of the forest inventory data and confirmed that it conforms to the level required by the project. The results of the field survey on utilization of forest biomass will be combined with the baseline biomass information to come up with the biomass potential for the Biomass Atlas of the forested areas. However, to date the open data-sharing policy has not been implemented, and access to the NFIS data has not been possible. Field survey on the crop biomass residues The field survey for crop biomass residue production was conducted by VNUA (based in Hanoi) and NLU (based in Ho Chi Minh City) (hereinafter referred to as "the Local Consultants") after being contracted by MOIT. The FA Consortium was responsible for monitoring and validating the said survey. The field survey was a person-to-person interview by the survey team with the farmers. It was executed with mobile phones using the MHG Mobile Application (a proprietary-software developed by the Consortium partner - MHG Systems Oy Ltd.). The phone application was used to record the responses of the interviewed farmer, indicate the location of the interview, and take geo- referenced photos of the surveyed crop fields. The detailed assignment of the Local Consultants and the procedure of conducting the field survey were described in the TOR for the Local Consultants, which were developed during Phase 1 of the project. The field survey on the crop biomass residues was carried out in 63 provinces and cities of Vietnam. A total number of 18,900 interviews was proposed (calculated as 63 provinces/cities x 5 main crops x 60 interviews/province). The exact number and locations of the survey/interviews (i.e., district and commune) per province or city were defined by the Local Consultants based on its preliminary Page | 15 discussion with the Department of Agriculture and Rural Development (DARD) of each province/city, and were submitted for approval to MOIT by the Local Consultants before executing the survey. The FA Consortium developed a complete crop biomass survey form (see section 4.1). The Local Consultants were trained by the FA Consortium on how to fill in the survey form using the smartphones during the Training Workshop on Field Survey and Data Collection organized on September 28-29, 2016 at NLU. 5.1.2 Validation of Crop Biomass Survey During the field interviews, Simosol downloaded the interview data from the MHG Mobile server, validated the data for logical consistency against agreed validation rules, and reported any invalid data back to the survey teams for checking as daily Excel survey data files, which highlighted the invalid data. The survey teams compiled the corrected daily Excel files into weekly files and sent those directly to Simosol for validation. Once the weekly file passed the validation rules, it was considered the final survey result file for that week. Another verification executed at Simosol was for the "reference field" of the survey questionnaire. The interviewed farmer selected a single field for which the crops cultivated between Nov '15 and Nov '16 were recorded and a photograph of the field was taken. If the crop information recorded and the photograph taken did not match, or the location of the photograph could not reliably be assigned to a field identifiable from a very high-resolution satellite image, the reference field was excluded from the land use classification reference dataset. Figure 6 shows a sample of an accepted reference field, and also demonstrates how the area to include into the reference sample was digitized during the validation process. Figure 6: A reference field sample that was included into land use classification reference sample data set. The yellow dots and white line delineate the area that was used as the reference sample for this field. Figure 7 gives an example of a reference field that was excluded from the reference data set. Page | 16 Figure 7: An example of a rejected reference field sample due to having been recorded in the middle of a road leaving uncertainty for the actual location of the field. 5.1.3 Results of Crop Biomass Survey VNUA completed their part of the crop biomass survey on December 8, 2016 and NLU on June 22, 2017. A total of 21,179 farmers in 514 districts were interviewed of which 13,835 interviews were conducted by NLU and 7,344 interviews by VNUA. After the validation, a total of 19,950 datasets were accepted of which 13,516 datasets were collected by NLU and 6,434 datasets were collected by VNUA. The number of surveyed districts, the number of farmers interviewed, and number of datasets accepted for each of the provinces are presented in Table 2. Table 2: Summary of number of districts and farmers surveyed Number of Number of Number of No. Province districts farmers datasets surveyed interviewed accepted 1 An Giang 8 515 515 2 Ba Ria-Vung Tau 5 213 213 3 Bac Giang 10 272 272 4 Bac Kan 4 110 74 5 Bac Lieu 4 292 269 6 Bac Ninh 8 83 83 7 Ben Tre 4 275 277 8 Binh Dinh 7 297 297 9 Binh Duong 7 369 373 10 Binh Phuoc 8 867 773 11 Binh Thuan 7 731 727 Page | 17 12 Ca Mau 4 273 274 13 Can Tho city 3 146 146 14 Cao Bang 12 238 172 15 Da Nang City 1 14 14 16 Dak Lak 12 999 987 17 Dak Nong 6 611 613 18 Dien Bien 7 279 280 19 Dong Nai 11 534 537 20 Dong Thap 6 489 489 21 Gia Lai 14 1,281 1,126 22 Ha Giang 11 387 309 23 Ha Nam 5 84 84 24 Ha Noi City 5 59 59 25 Ha Tay (now, Ha Noi City) 12 217 217 26 Ha Tinh 11 265 239 27 Hai Duong 12 171 171 28 Hai Phong City 2 96 96 29 Hau Giang 4 250 250 30 Ho Chi Minh City 6 132 125 31 Hoa Binh 11 255 201 32 Hung Yen 10 101 101 33 Khanh Hoa 5 174 278 34 Kien Giang 9 856 856 35 Kon Tum 8 401 402 36 Lai Chau 6 193 173 37 Lam Dong 10 726 583 38 Lang Son 11 210 210 39 Lao Cai 9 166 163 40 Long An 11 588 588 41 Nam Dinh 10 212 212 42 Nghe An 18 576 520 43 Ninh Binh 7 151 116 44 Ninh Thuan 4 148 148 45 Phu Tho 12 202 202 46 Phu Yen 6 245 245 47 Quang Binh 7 198 153 48 Quang Nam 8 224 224 49 Quang Ngai 5 268 268 50 Quang Ninh 9 93 94 51 Quang Tri 8 185 181 52 Soc Trang 7 405 405 53 Son La 11 634 541 54 Tay Ninh 9 625 625 55 Thai Binh 8 174 174 56 Thai Nguyen 7 401 210 57 Thanh Hoa 22 594 492 58 Thua Thien - Hue 9 114 114 59 Tien Giang 7 412 412 60 Tra Vinh 5 294 294 61 Tuyen Quang 6 188 160 62 Vinh Long 5 204 205 63 Vinh Phuc 9 140 105 64 Yen Bai 9 273 234 Total 514 21,179 19,950 Page | 18 Figure 8: Locations of farms surveyed Page | 19 Figure 9: Locations of farms with collected datasets accepted Page | 20 5.1.4 Lessons Learned from the Crop Biomass Survey Based on the report from VNUA, the following lessons were learned from the crop biomass survey:  The training of enumerators is extremely important. It should be done comprehensively.  Each trained enumerator should conduct at least five test surveys to make sure that he/she is fluent and familiar with the Survey App on smartphone, that he/she clearly understands the questionnaires, and develops interview skills.  The survey teams should have backup enumerators in case of emergency to assure the progress of the survey.  Each survey group (consisting of three survey teams) should have at least one local guide/contact person who knows the surveyed region (it can be an officer of DARD) to accompany the survey teams during the survey at the districts.  For remote areas where people do not speak Kinh (official Vietnamese) language, a translator is needed.  If possible, the Survey App should be designed in the way that invalid data are indicated immediately at the field so that enumerators can correct the data right away in order to save time and cost of redoing the survey.  Tablets (instead of mobile phone) are recommended for the survey due to longer battery autonomy and larger screen.  Avoid conducting the survey during the rainy season. 5.2 Collection and Validation of Industrial Biomass Data NLU was assigned to carry out the industrial biomass survey in the southern provinces of Vietnam in seven industrial sectors (including sugar mills, rice mills, MSW landfills, livestock farms, wood processing mills, brick-making factories, and pulp and paper mills). The purpose of the activity was to survey industrial sites which are producing secondary crop residues as part of their industrial processing chain, and/or utilizing biomass based feedstock for energy production and for other non-energy purposes. The type and amount of secondary crop residues produced would be mapped together with the amount that is actually used and the energy conversion technology. For the industrial sites using biomass, the type and amount of biomass used and the energy conversion technology in use would also be mapped. The GIS coordinates of the industrial sites are essential as the datasets without coordinates cannot be used for mapping. 5.2.1 Methodology of Industrial Biomass Survey With the support from FA Consortium, NLU has developed a methodology for the industrial biomass survey, including the six following steps:  Identification and preparation of a list of industries to be surveyed;  Sending of the survey questionnaires to the identified industries;  Compilation and analysis/validation of the data received; Page | 21  Selection of sites for on-site visits;  Conducting site visits to the selected industries;  Preparation of report on industrial biomass survey. NLU identified the major industries to be surveyed through the list of industries provided by provincial authorities (e.g. DARDs, DOITs, DONREs) and the relevant industrial associations (e.g. VSSA, AHAV, VDA), and from other publicly available sources. The background information on the industrial sites (such as name of industrial site, detailed address, email and telephone number of contact person, size of the industrial site, etc.) were also collected. The survey questionnaires developed by FA Consortium (see Attachment 2) were sent to the identified industries for collection of relevant data. Follow-up calls and e-mails were used by NLU to explain to the industries the purpose of the survey as well as the details of the survey questionnaire. The data received from the survey questionnaires were compiled and analyzed by NLU. Then, they were sent to FA Consortium for preliminary validation. Based on the analysis of the feedback from industries, and on the results of the preliminary data validation by FA Consortium, NLU selected sites for on-site visits to confirm the data provided in the complete questionnaire and, if needed, to collect additional information. 5.2.2 Validation of Industrial Biomass Survey FA Consortium carried out the validation of the data collected through industrial biomass survey by NLU. The preliminary validation was conducted for the preliminary datasets obtained through the survey questionnaires, and the final validation was done for the final datasets collected after on-site visits. NLU compiled the final datasets in an Excel file and sent it to FA Consortium for validation. The following key data/aspects were checked:  GIS coordinates of the surveyed sites (using Google map);  Residue-to-Crop Ratio (RCR);  Characteristics of the biomass residues (e.g. moisture content, LHV);  Completeness and consistency of the data. The checked datasets with errors/missing information highlighted were sent back to NLU. NLU re- contacted the contact persons at the surveyed industries to verify the incorrect data and/or to complete the missing information. The validated and verified data were eventually sent back to FA Consortium. That process continued until all datasets were finally accepted. 5.2.3 Results of Industrial Biomass Survey A total of 136 datasets from six industrial sectors were validated and accepted, of which 112 datasets were collected by NLU and 24 datasets were computed by FA Consortium from its database. Summary of the accepted datasets by industrial sector is given in Table 3. Page | 22 Table 3: Summary of the accepted datasets by industrial sector Datasets collected Datasets collected Sector Total by NLU by FA Consortium Sugar mills 3 13 16 Rice mills 46 0 46 MSW landfills 4 6 10 Livestock farms 14 5 19 Wood processing mills 29 0 29 Brick-making factories 16 0 16 Total 112 24 136 The industrial biomass survey was conducted in 16 cities/provinces in the South of Vietnam and in 2 other provinces in the central region. Only one pulp and paper mill in the South of Vietnam is using biomass residue for energy production. However, the survey for that mill could not be conducted as it was shut down for major maintenance. These datasets are provided in Attachment 4. The map of surveyed industrial sites is presented in Figure 10. Figure 10: Map of the surveyed industrial sites Page | 23 5.2.3 Lessons Learned from the Industrial Biomass Survey Based on the report from NLU, the following lessons were learned from the industrial biomass survey:  Difficulties were met in arranging meetings with industrial enterprises: no answer to request for meeting appointment, unable to meet managers;  Answerers/respondents from surveyed industrial enterprises do not necessarily understand the technical questions in the survey questionnaires;  Industrial enterprises do not archive their operational data through the years (especially, rice mills);  Some mills/factories did not keep any record of the type of generated biomass residues, and their quantities (especially, wood processing mills);  It is needed to closely coordinate with local authorities (in particular DOITs and DARDs) as their assistance is essential for successfully implementing the survey;  Before carrying out a survey in an enterprise, an email should be sent or a phone call given to clearly explain the purpose of the survey as well as the survey questionnaire;  There is a need to clearly explain the technical terms to the surveyed enterprises. 6. GIS DATA OF OTHER DRIVING COMPONENTS Currently, auxiliary GIS data used in the analysis consist of: 1. Power grid sub-station locations digitized at the World Bank from a paper map. 2. OpenStreetMap road and waterway network data. Currently, the World Band is sourcing topographic data including hydro/road network/admin boundaries/population/conversation areas and land use information. This data provided by MONRE will cover the 29 southern-most provinces from MONRE. In parallel MOIT is expected to acquire the data for the northern provinces. Should these datasets be available for the project during Phase 3, they can be incorporated in the data analysis for the final version of the Biomass Atlas. The same applies for the forest inventory data that has been requested from VNFOREST. 7. DATA ANALYSIS AND DEVELOPMENT OF THE DRAFT BIOMASS ATLAS 7.1 Potential Maps of Crop Harvesting Residues and Power Plant Site Suitability The first stage of development of the draft biomass atlas is illustrated in Figure 11. The first step of the analysis was the creation of the land cover classification. The details for this step are given in chapter 3.2 of this document. Page | 24 Figure 11: Components of the first atlas and harvest residue feedstock available at farm level The field survey data were used to derive the harvest residue feedstock map from the land cover classification map. The results for the survey were aggregated at district level. For each crop being analyzed, the crop yield and current utilization patterns for the harvest residues were aggregated from single responses to district level average values, as well as their standard deviation and min-max values. The utilization patterns modeling used the answers to these survey questions:  Residue use; used as fertilizer in the field, %  Residue use; sold to biomass trader, %  Residue use; other, %  Residue use; left in field for burning, %  Residue use; directly sold to industry, %  Residue use; as fuel in household, %  Residue use; animal fodder, % In addition, a set of questions about attitudes towards participation in a commercial feedstock supply chain were asked. In the analysis, only the portion currently left in field for burning was considered to be available as feedstock for power generation. The crop yield values from the survey responses were converted to different types of biomass residues based on residue-to-crop-ratios (see Table 4). Page | 25 Table 4: The residue-to-crop ratio (RCR) used in the analysis RCR Value Crop Residue used in Min Max Unit this study Crop harvesting biomass residues: Paddy Rice straw 1.00 0.33 2.15 t/t of paddy Maize Maize trash 2.20 1.00 3.77 t/t of maize (grain) Sugarcane Sugarcane trash 0.10 0.05 0.30 t/t of sugarcane (stem) Cotton Cotton stalks 3.40 2.76 4.25 t/t of cotton harvested Peanut Peanut straw 2.00 2.00 2.30 t/t of peanut (in-shell) Soybean Soybean straw 0.30 t/t of soybean (grain) Sweet potato Sweet potato straw 0.30 t/t of sweet potato Cassava Cassava stalks 0.30 0.06 0.30 t/t of cassava (manioc) Cashew nut Firewood 0.70 t/ha.year Rubber Firewood 0.50 t/ha.year Coffee Firewood 0.70 t/ha.year Tea Firewood 0.50 t/ha.year Pepper Firewood 0.50 t/ha.year Coconut Firewood 6.50 t/ha.year Grape Firewood 0.50 t/ha.year Mango Firewood 0.50 t/ha.year Orange Firewood 0.50 t/ha.year Mandarin Firewood 0.50 t/ha.year Longan Firewood 0.50 t/ha.year Litchi Firewood 0.50 t/ha.year Rambutan Firewood 0.50 t/ha.year Agro-industrial biomass residues: Paddy Rice husk 0.20 0.15 0.36 t/t of paddy milled Maize Corn cobs 0.30 0.20 0.50 t/t of maize (manioc) Maize Maize shells (husk) 0.20 0.20 0.40 t/t of maize (manioc) Sugarcane Sugarcane bagasse 0.30 0.14 0.40 t/t of sugarcane (stem) Peanut Peanut shells 0.30 0.30 0.48 t/t of peanut (in-shell) Cassava Cassava peels 0.12 0.10 0.15 t/t of cassava (root) Cashew nut Cashew nut shells 0.60 0.50 0.70 t/t of cashew nut (in-shell) Coffee Coffee husk 0.40 0.21 0.46 t/t of coffee bean Coconut Coconut husk 0.30 0.30 0.53 t/t of coconut fruit Coconut Coconut shells 0.15 0.12 0.15 t/t of coconut fruit Wood processing residues: Wood logs Wood edges, slabs, etc. 0.78 0.62 0.83 t/t of input logs Wood logs Sawdust 0.22 0.17 0.38 t/t of input logs MSW: MSW MSW combustibles 0.86 0.72 0.95 t/t of MSW MSW MSW organics 0.50 0.50 0.77 t/t of MSW Notes: Firewood refers to the tree bark, leaves and branches, shrubs, etc. from pruning the perennial industrial plants (tea, coffee, rubber, cashew, etc.), fruit plants (orange, lemon, longan, etc.). The biomass potential map was produced for each of the crop harvesting residues listed in Table 4. These maps show the amount of the crop harvesting residues available at the farms (see Figure 12). At the finest level of detail, these maps are at 20x20 m resolution, but for usability, they are further aggregated to 1,000x1,000 m resolution, which will reduce the file size to a usable level. Page | 26 Figure 12: Draft total technical harvest residue feedstock estimate map The second step of the data analysis and atlas creation is illustrated in Figure 13. Page | 27 Figure 13: Steps to create the industrial scale power generation potential atlas The output of the second step is a series of maps highlighting areas of high potential for industrial scale power generation. Each map is for one specific energy conversion technology and plant size. The maps are created based on a concept of a suitability index: the higher the index value, the higher the potential of the site for industrial power generation with the given technology. The set of analyzed technologies and plant sizes is:  Grate furnace type steam boiler + steam turbine: 3 MW, 8 MW, 15 MW.  Bubbling Fluidized Bed Combustion steam boiler + steam turbine: 8 MW, 15 MW, 25 MW, 50 MW, 100 MW.  Circulating Fluidized Bed Combustion steam boiler + steam turbine: 15 MW, 25 MW, 50 MW, 100 MW.  Gasifier + syngas engine/turbine: 0.5 MW, 1.5 MW.  Anaerobic digester + biogas engine: 0.5 MW, 1.5 MW, 3 MW, 8 MW This means that the final output for this second type of maps will consist of 18 suitability index maps. To derive these maps, we use the harvest residue feedstock maps to model the distance from which the feedstock must be sourced to the power plant. In order to derive the distance, the applicability of each feedstock for the given technology was graded with a scoring system: 0 – not suitable, 1 – acceptable, 2 – recommended. The biomass-to-power technology suitability matrix used is given in Attachment 5. The feedstock sourcing distance-modeling principle is illustrated in Figure 14. The same principle applies also to the other factors used in modeling the site suitability index. First a 1,000 x 1,000 m grid is spanned over the whole country. Then for each grid cell (one marked in the figure with "X") the minimum distance from which the total feedstock amount needed to operate the power plant Page | 28 can be sourced is computed using the feedstock map. This is done for the recommended feedstock for the plant that is most abundantly available: the shorter the sourcing distance, the higher the site suitability index value for the grid cell under analysis. Figure 14: The modeling principle for the different site suitability factors. However, before computing the site suitability for single fuel availability, the direct, "as-the-crow- flies" minimum sourcing distance is converted to an approximation of the real sourcing distance with the help of road network density in the area. The road network density, shown in Figure 15, was estimated based on road network data downloaded from the OpenStreetMap8. 8 http://wiki.openstreetmap.org/wiki/Downloading_data Page | 29 Figure 15: Road and watercourse network data used in the analysis The second index, after the single fuel sourcing distance index, to be computed for creating the final site-suitability index, was the multi-fuel sourcing index. Again, the feedstock suitability was modeled, but now conditional on the primary feedstock, and its minimum share of the feedstock mix (Attachment 5). Based on this grading, we computed a multi-fuel sourcing distance index similarly to the single fuel sourcing distance index. Similar distance-based index was computed for distance to the nearest power transmission network grid station (Figure 16) Page | 30 Figure 16: Grid stations, and the computed grid station distance index. (Darker green indicating higher index values, 50 km maximum distance for the grid connection) Finally, the three distance indices (single fuel sourcing, fuel mix sourcing and grid station connection) were combined as a single site-suitability index by weighing the individual indices (0.6 feedstock procurement distance, 0.3 for grid connection distance and 0.1 for road network density). The result is the final site-suitability index for each analyzed technology and power plant size combination. Figure 17 shows an example for a 15 MW circulating fluidized bed combustion steam boiler power plant. The site-suitability index is useful for greenfield projects looking for the most relevant places for an investment for further analysis. Feasibility analyses for power plant installation on an existing Page | 31 industrial site, can utilize the components of the atlas data that are used to derive the site-suitability index map. How to do this in practice will be the focus of Phase 3 training workshops. Figure 17: Site-suitability index for 15 MW circulating fluidized bed combustion based power plant. Index value ranges from 0 (not suitable, blue/white) to 100 (highly suitable, dark red) Page | 32 7.3 Biomass-Based Electricity Production Potential at the Existing Biomass Producing Sites An analysis of the existing agro-industrial and biomass producing sites covered by the industrial survey was conducted to assess the potential of each site for implementing a biomass-based power generating or cogeneration plant. 7.3.1 Sugar mills Based on the industrial survey, a total of 1.626 million tonnes/year of bagasse is generated in the 16 surveyed sugar mills in the South of Vietnam. 100% of this amount of bagasse is used as fuel in cogeneration plants to produce electricity and low-pressure process steam for covering the energy demand of the sugar mills. Most existing cogeneration plants are equipped with low-pressure steam boilers and back-pressure steam turbines working between 18 and 38 bar. Only one mill has already installed a high-pressure (98 bar) steam boiler. The total installed power capacity for the 16 surveyed cogeneration plants is estimated at 122 MW. There is a large potential for implementing new high-pressure cogeneration plants using bagasse at the sugar mills. The potential is calculated based on the assumption that all existing low pressure back-pressure steam turbine-based cogeneration systems will be converted into high-pressure systems using extraction condensing steam turbines. It should be noted that the use of new extraction condensing steam turbine allows the high-pressure cogeneration system to run during the off-milling season by utilizing additional biomass feedstock sourced from the vicinity of the sugar mill. Although the investment costs to convert the existing low-pressure to high-pressure cogeneration systems are high9, the implementation of high-pressure cogeneration systems should be considered as a priority for sugar mills in order to optimize the use of bagasse for power generation. Sticking to old, inefficient and polluting low-pressure systems should not be an option anymore. The process of analysis of new high-pressure cogeneration systems at sugar mills is as follows: (1) Calculating the energy input from bagasse (GWhth/year): For each sugar mill, the energy input from bagasse to the new high-pressure cogeneration plant is calculated based on the annual amount of bagasse generated at the sugar mill and the LHV of bagasse (7.5 MJ/kg). The data on bagasse generation was obtained from the industrial survey. (2) Calculating the total energy output from a cogeneration plant: An overall cogeneration efficiency of 75% was conservatively assumed for the cogeneration system at each sugar mill. Based on this assumption and the energy input from bagasse, the total energy output (i.e., steam thermal energy for process and electricity generation) from a cogeneration plant can be calculated. 9 Based on the Consultant’s experience in the region, the investment costs are USD 0.9 - 1.2 million/MW of installed power capacity for the capacity range of 15 - 35 MW. For the small-scale systems (<10 MW), the investment costs may reach USD 1.5 million/MW. Page | 33 (3) Calculating the process steam consumed by the sugar mill (GWhth/year): In order to calculate the process steam consumption (in thermal energy unit) of the sugar mill, it was assumed that 480 kg of low pressure steam (at 2.5 bar and 130oC) is consumed for processing one tonne of sugarcane. It should be noted that the current average process steam consumption in the Vietnamese sugar mills is 0.58 kg/t of sugarcane processed. Therefore, energy efficiency measures need to be implemented in order to reduce the process steam consumption by 15-20%, to 0.48 t/t of sugarcane processed. (4) Sizing of the cogeneration plant: The gross electricity generation of the cogeneration plant is calculated by subtracting the thermal energy of process steam consumed by the sugar mill from its total energy output. Then, the rated gross power capacity is defined based on the plant capacity factor (PCF) of the cogeneration plant, which is assumed to be equal to the operating time of the sugar mill during the milling season. (5) Calculating the net electricity output (GWh/year): This value is calculated from the gross electricity generation and an assumed parasitic load (electricity own-consumption) of the cogeneration plant. This value is dependent on the power capacity of the cogeneration plant: 20% for cogeneration systems <5 MWe; 15% for 5-10 MWe; 12% for 10-20 MWe and 9.0% for >20 MWe. (6) Calculating the electricity export to the grid (GWh/year) in case only bagasse is used: The amount of electricity exported to the grid is calculated by subtracting the electricity consumed by the sugar mill from the net electricity output of the cogeneration plant. The electricity consumption of the sugar mill is calculated based on the current electricity amount (kWh) required for processing one tonne of sugarcane. (7) Calculating the energy input from additional biomass feedstock (GWhth/year) for year-round operation of the cogeneration plant: Assuming that the annual PCF of the cogeneration plant increases to 85%, the annual PCF of the cogeneration plant running on additional biomass feedstock during the off-milling season is calculated. Based on this PCF value and the rated gross power capacity defined in step (4), the gross electricity generation of the cogeneration plant running on additional feedstock can be calculated. Then, the amount of energy input from additional biomass feedstock is calculated using an assumed value of 25% for electrical efficiency of the cogeneration plant running in pure-power generation mode. For the additional biomass feedstock, the fuel deterioration during storage for the off-milling season was taken into account. (8) Calculating the total electricity export to the grid (GWh/year): The net electricity output of the cogeneration plant running on additional biomass feedstock is calculated from the gross electricity generation defined in step (7) and the plant parasitic load. (9) Calculating the amount of additional biomass feedstock (tonnes/year) and its sourcing area (km²/GWh): This is done based on the amount of energy input calculated in step (7) and the map of technical potential of crop harvesting residues. In other words, the sourcing area for additional biomass feedstock takes into account the real distribution of the crop fields and crop residues within the vicinity of the mill. Page | 34 The additional biomass feedstock sourcing area matches the best-case scenario, which is able to source all of the technically available crop harvesting residues suitable for the cogeneration plant from the immediate neighborhood of the sugar mill. Therefore, it helps ranking the sugar mills in terms of the ease of sourcing the additional biomass feedstock. The results of the sugar mills analysis show that the new high-pressure cogeneration plants at 16 sugar mills could have a combined power capacity output of around 205 MW which is about 1.7 times higher than the total power capacity of all 16 existing low-pressure cogeneration plants. In order to run these cogeneration plants at an annual PCF of 85%, around 1.806 million tonnes/year of additional biomass feedstock are needed. A total amount of about 330 GWh/year of electricity could be exported to the grid if only bagasse is used during the milling season or 1,174 GWh/year if both bagasse and additional biomass feedstock are used as fuels for the cogeneration plants all year round. Table 5 summarizes the results of the analysis of the 16 surveyed sugar mills. The detailed results for each sugar mill are provided in Attachment 6. Table 5: Summarized results of analysis of 16 sugar mills Description Unit Value Number of mills 16 Total sugarcane crushing capacity t/day 49,950 Total sugarcane processed t/year 5,361,000 Total bagasse generated t/year 1,626,000 Total additional biomass feedstock sourced t/year 1,806,000 Power generation technology used High pressure Total gross power capacity output MW 205 Total gross electricity generated GWh/year 1,528 Total net electricity output GWh/year 1,362 Total electricity consumption by the sugar mills GWh/year 188 Total electricity exported to the grid, of which: GWh/year 1,174 From bagasse only (during the milling season) GWh/year 330 From additional feedstock (during the off-milling season) GWh/year 844 The map of potential high-pressure cogeneration plants at the 16 surveyed sugar mills in the South of Vietnam is shown in Figure 18. In this figure, the size of the circle is proportional to the potential power plant capacity (ranging from 3.7 MW to 45 MW), and the color of the circle relates to the sourcing area for additional biomass feedstock (km2 for each additional GWh required). The sourcing areas range from dark red at 0.13-0.57 km²/GWh to dark blue at 1.92-2.36 km²/GWh, and other color hues in between those two ranges. Page | 35 Figure 18: Map of potential high-pressure cogeneration plants at the existing sugar mills in the South of Vietnam 7.3.2 Rice mills It must be mentioned that only 46 rice mills provided adequate data during the industrial survey (see Attachment 4). The total amount of paddy milled in these 46 surveyed rice mills was 4.216 million tonnes/year which accounted for only 9.6% of total amount of paddy produced in Vietnam in 2016 (43.69 million tonnes). The total amount of rice husk generated from these 46 rice mills was about 853,460 tonnes/year of which 238,340 tonnes/year (28% of the total) was used by the rice mills for in-house purposes (i.e., drying paddy, producing rice husk pellets, etc.). The surplus amount of rice husk (615,120 tonnes/year) was being sold out. If this amount would be used for power generation by the rice mills, it could support around 66 MW of power capacity, i.e. an average of around 1.4 MW per mill. As this potential is too low to attract investors, the use of additional biomass feedstock was considered in the analysis of the power generation potential in order to increase the capacity of each power plant to be able to export power to the grid. It was assumed that the steam boiler of the power plant would be run on rice husk or on other locally sourced biomass feedstock, or on a mixture of them. The minimum fixed power plant capacity of 3 MW is assumed for the power plants at all 46 surveyed rice mills. The additional biomass feedstock was calculated in order to assure an annual plant capacity factor of 80% for all power plants. As for the sugar mills, the analysis results for each rice mill contain the sourcing area (km²/GWh) for the additional biomass feedstock needed to operate the power plant, with the sourcing area matching the best-case sourcing scenario. Page | 36 The process of analysis of potential power plants at rice mills is as follows: (1) Calculating the energy input from rice husk (MWhth/year): The energy input from rice husk is calculated based on the surplus rice husk of each rice mill (obtained from the industrial survey) and its LHV of 13.0 MJ/kg. (2) Calculating the gross power capacity output of the power plant (MW): The gross power capacity output of the power plant is calculated based on the energy input from rice husk, the electrical efficiency of the power plant (medium-pressure steam-cycle technology with an average gross electrical efficiency of 20.8% is used), and on annual PCF of 80%. If the calculated gross power capacity output is less than 3 MW, such power plant is assumed to have a fixed power capacity of 3 MW. (3) Calculating the gross electricity output of the power plant (GWh/year): The gross electricity output of the power plant is calculated based on its gross power capacity output and operating time (at 80% annual PCF, the operating time of the power plant is 7,008 hours/year). (4) Calculating the net electricity output (GWh/year): This value is calculated from the gross electricity output and an assumed parasitic load (electricity own-consumption) of the power plant. This value is dependent on the power capacity of the power plant: 15% for power plants <5 MWe; 12% for 5-10 MWe; 10% for 10-20 MWe. (5) Calculating the electricity export (GWh/year): The amount of electricity exported to the grid is calculated by subtracting the electricity consumed by the rice mill from the net electricity output of the power plant. The electricity consumption of the rice mill is calculated based on the current electricity amount (kWh) required for processing one tonne of paddy. (6) Calculating the energy input from additional biomass feedstock (MWhth/year): For a power plant where the amount of surplus rice husk is not sufficient for generating 3 MW gross power capacity, additional feedstock has to be sourced. The energy input from additional feedstock is calculated by subtracting the amount of energy input from rice husk from the total amount of energy input required for a 3 MW power plant (101,077 MWhth/year). (7) Calculating the amount of additional biomass feedstock (tonnes/year) and its sourcing area (km²/GWh): This is done based on the amount of energy input calculated in step (6) and the map of technical potential of crop harvesting residues. It takes into consideration the real distribution of fields and crops, but assuming 100% sourcing ability of the available feedstock. The effect of fuel deterioration during the storage period was also taken into account. The results of the rice mills analysis show that, the calculated combined power capacity output of the potential power plants at 46 rice mills is 164 MW. A calculated amount of biomass fuel of 1.623 million tonnes/year is required of which 0.615 million tonnes/year come from rice husk and 1.008 million tonnes/year from additional biomass feedstock. These potential power plants could export about 863 GWh/year of electricity to the grid. Table 6 summarizes the results of the analysis of the 46 surveyed rice mills. The detailed results for each rice mill are provided in Attachment 6. Page | 37 Table 6: Summarized results of analysis of 46 rice mills Description Unit Value Number of mills 46 Total paddy milling capacity t/day 30,290 Total paddy processed t/year 4,215,560 Total rice husk generated t/year 853,460 Total rice husk used by the rice mills t/year 238,340 Total surplus rice husk (used for power generation) t/year 615,120 Total additional biomass feedstock sourced t/year 1,007,950 Power generation technology used Medium pressure Total gross power capacity output MW 164 Total gross electricity generated GWh/year 1,153 Total net electricity output GWh/year 989 Total electricity consumption by the rice mills GWh/year 126 Total electricity exported to the grid, of which: GWh/year 863 From rice husk only GWh/year 346 From additional feedstock GWh/year 517 The map of potential power plants at the surveyed rice mills in the South of Vietnam is provided in Figure 19. The bigger the circle, the higher the capacity of the power plant (ranging from 3 MW to 9 MW). The red circles show “zero” sourcing area for the additional feedstock (i.e., no additional feedstock needed), the dark blue circles indicate sourcing area ranging from 0.36 to 0.47 km2/GWh while other colors are in between. Figure 19: Map of potential power plants at 46 existing rice mills in the South of Vietnam Page | 38 7.3.3 MSW landfills There is a potential use of MSW for electricity generation at the landfills in Vietnam. Two technologies which are most frequently used for generating electricity from MSW include:  direct combustion of organic materials of MSW in an incinerator/steam boiler to produce high pressure steam. Then, this steam is used in a steam turbo-generator to generate electricity;  anaerobic digestion of biodegradable fraction of MSW to produce biogas which is used in a gas engine/turbine system to generate electricity. The use of incinerator/steam boiler to convert MSW to electricity is a relatively old technique. The electrical efficiency of this technology is low (typically, 14-28%) due to the low efficiency of the incinerator/steam boiler burning MSW with high moisture content and low LHV. Another problem associated with direct combustion of MSW to generate electricity is the potential for pollutants (such as nitrogen oxides, sulphur dioxide, heavy metals and dioxins) to enter the atmosphere with the flue gases from the incinerator/boiler. In this study, the anaerobic digestion of MSW is proposed as it has a higher electrical efficiency and a lower environmental impact than the direct combustion technology. The following steps were used for analyzing the potential power plants installed at the landfills based on anaerobic digester combined with gas engine/turbine systems: (1) Calculating the amount of organic MSW for each landfill (t/year): This amount is calculated based on the amount of MSW disposed at each landfill (obtained from the industrial survey) and the percentage of organic matters in MSW (an average value of 50% is assumed for the MSW landfills in Vietnam). (2) Calculating the annual biogas production for each landfill (m3/year): Based on the amount of organic MSW at each landfill, the annual biogas production is calculated by using an average biogas production rate of 120 m3/tonne of organic MSW (as received basis). (3) Calculating the energy input from the biogas to the gas engine/turbine (GWhth/year): This value is calculated based on the annual biogas production and biogas LHV. Assuming that biogas produced from MSW consists of 60% of methane, the biogas LHV was calculated at 21.54 MJ/m3. (4) Calculating the gross electricity output from the power plant (GWh/year): The gross electricity output was calculated based on an assumed gross electrical efficiency of 40% for biogas engine-based technology and 30% for biogas turbine-based technology. (5) Calculating the rated gross power capacity (MW): The rated gross power capacity of each power plant was calculated based on its gross electricity output and an annual PCF of 90%. (6) Calculating the electricity export (GWh/year): The amount of electricity export was calculated based on an average parasitic load of 5% assumed for all power plants. Page | 39 The industrial survey covered 10 landfills in the South of Vietnam. The combined amount of MSW collected at these landfills is around 9,300 t/day. However, some amount of MSW (605 t/day) is being used for fertilizer production, and the remaining amount of around 8,695 t/day is currently dumped at the 10 landfills. This amount of MSW generates around 4,347 t/day of organic MSW. If this organic MSW could be used for electricity generation, around 58 MW of gross power capacity could be generated in the anaerobic digester-based power plants. Table 7 summarizes the results of analyzing 10 surveyed MSW landfills. The detailed results for each landfill are provided in Attachment 6. Table 7: Summarized results of analysis of 10 MSW landfills Description Unit Value Number of landfills 10 Total MSW collected t/day 9,300 Total MSW dumped t/day 8,695 Total organic MSW generated t/day 4,347 Total annual biogas production million Nm3/yr 190.4 Power generation technology used Biogas engine Total gross power capacity output MW 58 Total gross electricity generated GWh/year 455 Total net electricity output GWh/year 433 Total electricity exported to the grid GWh/year 433 The map of potential power plants at the surveyed MSW landfills in the South of Vietnam is provided in Figure 20. The bigger the circle, the higher the capacity of the power plant (ranging from 0.8 MW to 33 MW). Figure 20: Map of potential power plants at the surveyed MSW landfills in the South of Vietnam Page | 40 7.3.4 Livestock farms The following steps were used for analyzing the potential power plants installed at the livestock farms, based on an anaerobic digester combined with gas engine/turbine systems: (1) Calculating the annual biogas production for each livestock farm (m3/year): The annual biogas production is calculated based on the amount of manure available at each livestock farm (obtained from the industrial survey) and the average biogas production rate (m3/t of manure on an as received basis). Based on the industrial survey, the biogas production rate is 0.030 m3/t for cow manure, and 0.040-0.057 m3/t for pig manure. (2) Calculating the energy input from biogas to the gas engine/turbine (GWhth/year): This value is calculated based on the annual biogas production and biogas LHV. Assuming that biogas produced from livestock manure consists of 65% of methane, the biogas LHV was calculated at 23.34 MJ/m3. (3) Calculating the gross electricity output from the power plant (GWh/year): The gross electricity output was calculated based on an assumed gross electrical efficiency of 40% for biogas engine-based technology and 30% for biogas turbine-based technology. (4) Calculating the rated gross power capacity (MW): The rated gross power capacity of each power plant was calculated based on its gross electricity output and an annual PCF of 90%. (5) Calculating the electricity export (GWh/year): The amount of electricity exported to the grid was calculated based on an average parasitic load of 5% assumed for all power plants. The industrial survey covered 19 livestock farms in the South of Vietnam, including 4 cow farms and 15 pig farms. The combined amount of manure collected at these farm is around 721.6 t/day. Around 43% of the collected manure (311.2 t/day) is being used for fertilizer production (4 farms are using all 100% of their manure for that purpose). The remaining amount of 411.4 t/day (150,163 t/year) of manure available at 15 livestock farms is currently used for production of 15,380 m3/day (5,613,700 m3/year) of biogas. If this volume of biogas could be used for electricity generation, around 1.846 MW of gross power capacity could be generated in the anaerobic digester-based power plants. Table 8 summarizes the results of the analysis of the 19 surveyed livestock farms. The detailed results for each farm are provided in Attachment 6. Table 8: Summarized results of analysis of 19 livestock farms Description Unit Value Number of livestock farms 19 Total manure collected t/day 722 Total manure used for fertilizer production t/day 311 Total manure used for biogas production t/day 411 Total volume of biogas produced m3/year 5,613,700 Power generation technology used Biogas engine Total gross power capacity output MW 1.846 Total gross electricity generated GWh/year 14.55 Total net electricity output GWh/year 13.82 Total electricity exported to the grid GWh/year 13.82 Page | 41 The map of potential power plants at the surveyed livestock farms in the South of Vietnam is provided in Figure 21. The bigger the circle is, the higher the capacity of the power plant (from 4 to 614 kW). Figure 21: Map of potential power plants at the surveyed livestock farms in the South of Vietnam 7.3.5 Wood Processing Mills The industrial survey covered 29 wood processing mills in the South of Vietnam (see Attachment 4). The total amount of wood logs processed in these mills was 167,710 tonnes/year. The total amount of wood residues (sawdust, edges and slabs) generated from these 29 surveyed wood processing mills was about 78,250 tonnes/year of which 41,250 tonnes/year (52.7% of the total) was used by the mills for in-house purposes (mainly for wood drying). The surplus amount of wood residues (37,000 tonnes/year) was being sold. If this amount would be used for power generation by the wood processing mills, it could support around 4.2 MW of gross power capacity. This potential is small. The detailed analysis of the electricity potential from wood processing mills considering additional biomass feedstock will be carried out and reported in Phase 3. 8. DATA VALIDATION WORKSHOP A Data Validation Workshop was held on 15 November 2017 at MOIT (23 Ngo Quyen street, Hoan Kiem District, Hanoi City, Vietnam). The main objective of this workshop was to present the draft Biomass Atlas to the stakeholders for validation. It aimed at seeking feedback from the stakeholders Page | 42 on the data collected and on the draft Biomass Atlas in order to finalize and produce a high-quality validated Biomass Atlas for Vietnam. The topics of this multi-stakeholder workshop included the following:  Progress updates of the Biomass Resource Mapping Project for Vietnam;  Presentation of the crop biomass survey, including the survey methodology, results and lessons learned;  Presentation of the industrial biomass survey, including the survey methodology, results and lessons learned;  Presentation of the mapping methodology and key limitations of the Biomass Atlas for Vietnam;  Presentation of the draft Biomass Atlas for Vietnam, including the potential maps of crop harvesting residues and power plant site suitability and the potential for industrial scale biomass- based electricity production at the existing industrial sites. A total of 24 participants attended the Data Validation Workshop. The participants consisted of representatives from WB, MOIT, the provincial Departments of Industry and Trade (DOITs), from central government authorities, universities and research institutions and the consulting team. The completion report on this workshop is provided in Attachment 7. 9. REVISED WORK PLAN FOR PHASE 3 In order to plan Phase 3 implementation, FA Consortium met with the WB (Mr. Tran Hong Ky) and MOIT (Ms. Pham Thuy Dung and Ms. Ngo To Nhien) representatives on 14 November 2017. Based on the discussions with the WB and MOIT representatives, the Work Plan for Phase 3 of the project was revised. It is presented in Table 9. Page | 43 Table 9: Revised Work Plan of Phase 3 2016 2017 2018 Date ACTIVITY Lead Support 8 9 10 11 12 01 02 03 04 05 06 07 08 09 10 11 12 01 02 03 04 05 06 07 08 Covered Phase 2 1 Task 2.1 Remote Sensing Data 01/08/16 SML Collection 30/08/17 Task 2.2 Support on-site data collection campaign 2.1 Preparation & execution of 28/09/16 SML, IE, FA 2 Training Workshop 29/09/16 ET 2.2a Validation of field data 01/10/16 SML, collected by local consultants IE, ET 30/05/17 FA (Field crop survey) 2.2b Validation of field data 01/06/17 SML, collected by local consultants IE, ET 30/08/17 FA (Industrial survey in the South) 3 Task 2.3 Data analysis & 01/09/17 SML, FA, WIL, production of Draft Biomass 31/10/17 VTT IE, ET Atlas 4 Task 2.4 One-day Validation 01/11/17 FA, IE, ET Workshop 30/11/17 SML 5 Deliver Phase 2 Outputs FA, 20/12/17 IE, ET (Implementation Report) SML Phase 3 2.2c Data validation of the 24/11/17 SML, industrial survey in the IE, ET 15/01/18 FA Northern & Central provinces 6 Task 3.1 Final analysis and 15/01/18 FA, IE, SML mapping 15/02/18 ET 7 Task 3.2 Production of final 15/02/18 FA, IE, SML Biomass Atlas 28/02/18 ET 8 Task 3.2 Dissemination 01/03/18 FA, workshop and training program IE, ET 31/03/18 SML (3 days) Page | 44 9 Deliver Phase 3 Outputs FA, 30/04/18 IE, ET (Draft Final Report) SML 10 Peer-review & comment by 01/05/18 WB the WB’s experts 31/07/18 11 Finalization and submission 01/08/18 FA, of the Final Report 31/08/18 SML LEGEND: Deliverables Milestones Home Work Field Work FA - Full Advantage Co., Ltd. (Thailand); SML - Simosol Oy (Finland); VTT - VTT Technical Research Center (Finland); WIL – Wiltrain Oy (Finland); IE – Institute of Energy (Vietnam); ET – Energy Conservation Research and Development Center (Vietnam) Page | 45 10. CONCLUSIONS AND RECOMMENDATIONS 10.1 Conclusions Remote Data Collection and Analysis: During Phase 3, the land use classification will be finalized by testing different approaches for improving the crop classification accuracy: addition of optical satellite images to the analysis, delineation of the analysis area using auxiliary GIS data sources, and different modelling approaches to reduce the over-estimation of dominant crop classes for each agro-ecological zone. Preparation for Field Survey and Data Collection: A training workshop on field survey and data collection of crop biomass residues in Vietnam was conducted at Nong Lam University in Ho Chi Minh City on 28-29 September 2016. It provided a hands-on training to the field surveyors from NLU AND VNUA for conducting surveys using 3G smartphones and Android applications. A smartphone based field survey questionnaire using the MHG Mobile app was developed for training and survey purpose. Crop Biomass Survey: The crop survey activities were conducted in collaboration with the two universities and the agriculture departments in the different regions of Vietnam using digital survey tools. The questionnaire was designed to explore the net availability of biomass or crop residue for power production, excluding the utilization of biomass as feed or at domestic and industrial level. The coordination between the international consultants and the two universities and further between the universities and the agriculture departments helped successfully complete the crop survey in Vietnam. Industrial Biomass Survey: The industrial survey was conducted by NLU in the southern provinces of Vietnam. The industrial survey should be conducted by VNUA in the northern and central provinces in December 2017 and January 2018. Data Analysis and Development of a Draft Biomass Atlas: The Biomass Atlas will contain two sections: the first one related to feedstock availability at farm level and the second one related to site-suitability. They are based on a set of modeling assumptions in addition to the land use classification and collected survey data. They are also subject to the inherent uncertainty in the crop yield levels and the field survey results. This means that the users of the Atlas must be aware of these assumptions and the degree of uncertainty. Delivering those together with the Atlas is essential for their proper use. This will also be highlighted in the Training Workshops to be arranged at the end of Phase 3. Another aspect to be focused on in the Training Workshops is how to utilize the atlas data for feasibility analysis of brownfield projects. That would focus on a single location, without being fully in line with the atlas concept. However, the data behind the atlas also fully support this kind of analysis. Page | 46 Data Validation Workshop: A data validation workshop was conducted in Hanoi on 15 November 2017 to present the draft Biomass Atlas to the stakeholders for validation and verification. The feedback from the stakeholders was received. They will be incorporated in the final processing of the datasets and production of the final Biomass Atlas for Vietnam. Phase 3 Work Plan Revision: The Work Plan for Phase 3 of the project was revised based on discussions with the WB and MOIT representatives and on the feedback of participants during the Data Validation Workshop. The production of the final Biomass Atlas and associated GIS files/datasets is planned to be completed by the end of February 2017. The Atlas Dissemination Seminar and Training Workshops will then be conducted during March 2017 after which the Final Report will be submitted to the WB/MOIT. 10.2 Recommendations It is important to take a decision on who will take over the Final Biomass Atlas and be responsible for updating it. This decision should be taken as soon as possible in order for the Consortium to appropriately plan and prepare the Atlas Maintenance Training Workshop of March 2017. The participants in the Biomass Atlas Usage and Maintenance Training Workshops should be carefully selected to assure the success of these trainings. The attendees of these training workshops should be fluent with IT and GIS. 11. ATTACHMENTS Attachment 1: Crop biomass survey form Attachment 2: Industrial biomass survey questionnaires Attachment 3: Report on training workshop on field survey and data collection Attachment 4: Report on industrial biomass survey Attachment 5: Biomass-to-power technology suitability matrix Attachment 6: Results of analysis of electricity production potential at industrial sites Attachment 7: Report on Data Validation Workshop Page | 47