Dniester River Reservoir Simulation Modeling Workshop Summary Report Contents Introduction .................................................................................................................................................. 1 Summary of Workshop Components ............................................................................................................ 2 Introduction to the Multi-Reservoir Model .............................................................................................. 2 Application to Planning ............................................................................................................................. 3 Application for Real-time .......................................................................................................................... 3 Participant Presentations.......................................................................................................................... 3 Basic Demonstration of Model ................................................................................................................. 4 Feedback from Participants .......................................................................................................................... 5 Ukraine .................................................................................................................................................. 5 Moldova ................................................................................................................................................ 7 Key Outcomes ............................................................................................................................................... 9 Future Work ................................................................................................................................................ 10 Introduction The United Nations Economic Commission for Europe (UNECE) and the Organization for Security and Cooperation in Europe (OSCE) are leading a project component on “Climate Change and Security in the Dniester River Basin,” which aims to improve transboundary water management and build the capacity to adapt to climate change within the Dniester River Basin. This effort is one of several of a larger project on “Climate Change and Security in Eastern Europe, Central Asia and the Southern Caucasus” implemented within the framework of the Environment and Security (ENVSEC) Initiative and funded through a contribution from the European Commission’s Instrument for Stability and with further cofunding from the Austrian Development Agency and the ENVSEC partners. Since July 2013, the U.S. Army Corps of Engineers (USACE) Institute for Water Resources (IWR) has been working on the project on behalf of the Alliance for Global Water Adaptation (AGWA). The goal of this activity is to develop a risk-informed decision framework for water resources adaptation to climate change. As part of this effort, IWR has developed a multi-reservoir operating model for the Dniester River basin, containing reservoirs in both Moldova and Ukraine, to evaluate climate adaptation strategies for reservoir management. The USACE team is developing two key products: (1) a working HEC-ResSim model that will integrate the flows and reservoir operations of the main reservoirs in Moldova and Ukraine and (2) a 1 framework for risk-based decision making and evaluation focused on reservoir management and future investment decisions needed for climate adaption in three key sectors – agriculture, flood control and environmental protection. The idea of application of such a model to the Dniester river basin was suggested by one of the project main stakeholders (Ukrainian Hydrometeorological Centre) in July 2013 and was discussed with all relevant members of the project Working Group on Adaptation to Climate Change and Flood Management. Later this measure was selected as one of the priorities for further implementation according to key adaptation needs in the basin, importance and impact of the reservoirs for the management of the basin and their ability to serve as an adaptation tool. The USACE team, in collaboration with UNECE, OSCE, AGWA, and the International Center for Integrated Water Resources Management under the auspices of UNESCO (ICIWaRM), held workshops in Kiev, Ukraine on 22-24 October and in Chisinau, Moldova on 28-30 October 2014. Workshops were technically organized and facilitated through the contract with the Global Water Partnership (Andriy Demydenko). The workshop objectives were to: (1) demonstrate climate change adaptation options through reservoir operations and transboundary water management; (2) introduce capabilities of the multi-reservoir operating model for planning and real-time forecasting; (3) provide a demonstration of the model so participants are able to use as their own resource; and (4) discuss future model applications and development. Participants at both workshops included representatives from the national authorities, academia, and non-governmental organizations. Summary of Workshop Components The workshop consisted of four components: (1) introduction to the multi-reservoir model; (2) presentation of planning applications; (3) presentation of real-time applications; and (4) hands-on demonstration of the HEC-ResSim software and the Dniester River basin model. Discussion sessions were scheduled throughout each of these workshop components to receive feedback from participants and adjustments to the agenda were made as needed. Introduction to the Multi-Reservoir Model The first portion of the workshop focused on introducing the model to workshop participants. Over the past six months, the USACE team has developed the multi-reservoir model for the Dniester River Basin, with assistance from in-country subject matter experts, using the USACE Hydraulic Engineering Center (HEC) Reservoir System Simulation (ResSim) software program. ResSim, a public domain software program, is used to assist with water resources planning and real-time operations. The software program simulates system behavior in terms of reservoir storage management, downstream flow control, and withdrawals. Current capabilities include release allocation schemes for reservoirs with multiple outlets and coordinated operation for a system of reservoirs. The model also simulates both incidental and scheduled hydropower production. In addition, operating constraints can be modeled using conditional statements and scriptable functions, which allow greater flexibility in dealing with varied basin conditions and more involved regulation plans. The Dniester River Basin ResSim model consists of a Dniester River network from Zaleschiky to the downstream Estuary in Ukraine. The reservoirs modeled include the Dniestrovsky, Buffer, and Pumped Storage in Ukraine and the Dubasari in Moldova. Diversions for agricultural, municipal, and industrial water supply purposes were located at Egoreni and Eerpeni (Moldova), and Odessa (Ukraine). Junctions 2 to evaluate flood flows are located at Mogyliv-Podilsky and Bender. Another junction is located at the estuary to evaluate flows and their potential environmental impacts. The input data for the model are observed flow data at Zaleschiky from 1947-2010 as well as local inflow information for tributaries throughout the watershed. Reservoir physical characteristics and operating constraints are based on information provided by subject matter experts, published journal articles, and existing reservoir manuals. Application to Planning Following an overview of the Dniester River Basin ResSim model, Drs. Rolf Olsen and Kristin Gilroy from the USACE team presented potential applications of the model for short term and long term planning purposes. The model was used to explore planning questions such as: • • • • Does the system meet demands under existing conditions? Can the system accommodate increased demand? What effect will climate change have on the system? What adaptation measures can be implemented? For short-term planning, trade-offs between operating for hydropower versus the environment were explored for observed drought conditions. For long-term planning, performance metrics for flood risk management, hydropower, water supply, and the environment were explored for (1) existing conditions over the entire period of record; (2) future conditions with a wetter climate; and (3) future conditions with increased demands and drier climate. Finally, an adaptation measure for flood risk management based on shifting the Dniestrovsky reservoir flood storage to the summer months to accommodate summer flooding was explored. The analyses were discussed in detail with the participants. Application for Real-time In addition to planning applications, Fauwaz Hanbali from the USACE team provided an overview of how the model can be used for real-time forecasting. Hanbali presented on the Corps Water Management System and Model Integration Software (HEC-RTS) available through USACE HEC. HEC-RTS is a suite of software for real time forecasting and water system management decision support, consisting of realtime data acquisition and validation and integration of meteorological, rainfall runoff, and reservoir and river system models. ResSim is the component within this system for reservoir system simulation. Following the overview of HEC-RTS, Hanbali lead a discussion of the Dniester River Basin’s existing data monitoring infrastructure and hydrologic and hydraulic modeling to support real-time forecasting. Participant Presentations At both workshops, local agencies and researchers were provided an opportunity to present their work regarding water management within the Dniester River basin. Victoria Boyko provided an overview of the existing capabilities of Ukrainian Hydrometeorological Centre and how the ResSim model could be used to assist in meeting their objectives. The following four areas were discussed: (1) mmonthly modeling; (2) ecological flow modeling; (3) flood modeling and flood frequency analyses; and (4) modeling of hydrographs. Each month the Interdepartmental Commission under the State Agency of Water Resources of Ukraine determines releases from the Dniester reservoirs. The Commission consists of representatives of water management, hydropower, fishery, environmental protection and other agencies. The Ukrainian Hydrometeorological Center 3 provides a forecast of the monthly inflow to the Dniester HES-1 (based on the current water flow and its retrospective analysis) that is used to set the release amounts. Since the early 1990s, artificial floods were conducted in April and May to improve the ecological balance and fish reproduction in the Dniester River downstream of the Dniestrovsky Dam including the Dniester delta. The ResSim model could be used to forecast the implications of the release. There is a 10 centimeter limit on daily fluctuations in the reservoir that affects the releases; however, during several last years the elevation change was higher due to the needs of the ecological release. In Ukraine, Mark Zheleznyak from the Ukrainian Centre of Environmental and Water Projects discussed two modeling tasks on climate change and flooding implemented within the preceding project “Reducing Vulnerability to Extreme Floods and Climate Change in the Dniester River Basin”. They developed a one-dimensional hydraulic model downstream of the buffer dam and a two-dimensional hydraulic model that showed flood inundation in Mogyliv-Podilsky. He estimated that with climate change the same frequency flood may be about 15% larger. In Moldova, Gherman Bejenaru from the State Hydrometeorological Service of the Republic of Moldova provided a presentation of ongoing flood and environmental modeling efforts in Moldova. The European Investment Bank is funding the development of a countrywide flood management Master Plan. BETA Studio srl and HR Wallingford Ltd are performing the work which is partly (mainly on the Dniester delta) supported by the following projects: “Climate Change and Security in the Dniester River Basin” and “Reducing Vulnerability to Extreme Floods and Climate Change in the Dniester River Basin”. Basic Demonstration of Model The USACE team provided a demonstration on development and use of the ResSim model. Hanbali showed how to develop a ResSim model, beginning with data needs and ending with a complete model constructed for the Dniester River basin. Olsen and Hanbali then worked with participants to install the ResSim software program and download the Dniester River Basin model on each participant’s computer. Gilroy then led interactive exercises demonstrating how to use the model to analyze the effect of different operating constraints on performance metrics for each water sector. Participants were able to make changes to operating constraints within the model and analyze results ranging from downstream flows to reservoir elevations to hydropower generation. Olsen provided a demonstration of the model’s hydropower capabilities and an overview of the multiple hydropower plants represented in the model. As a conclusion to the training, Hanbali provided an interactive demonstration of real-time operations using the 2008 flood. Participants suggested different operations to mitigate the flood based on varying forecast lead times until the system was able to pass the peak flow through the system without exceeding any flood risk constraints. Hanbali showed that if the initial reservoir level was 121 meters, the lead time for a flood forecast would need to be four days to have enough time to empty flood storage space to prevent flood damages downstream. Four days generally exceeds the lead time for accurate precipitation forecasts. At the end of this basic training, participants had a good understanding of the capabilities of the model and were able to analyze the results for different reservoir operations and climate and demand scenarios with the model. The time constraints on the workshop limited the material accessible for participants; however, the USACE Hydrologic Engineering Center (HEC) provides extensive training courses for interested parties. The potential for future trainings was discussed. 4 Feedback from Participants Ukraine Feedback on Dniester ResSim Model The participants made comments on the model and recommended some changes. There are no operating constraints on the maximum or minimum rate of change of the releases from Dniestrovsky dam. There is a constraint of 10 centimeters per day on the rate of change of the water level at the Dniestrovsky reservoir. The reservoir guide curve shown in Figure 1 is not used for regulatory purposes. The normal level for the Dniestrovsky reservoir is 121 meters. Reservoir managers will try to stay at this level if inflows support it; however, after the high flood in 2008 the reservoirs managers often keep the level at 119 meters to allow larger storage in case of flood. The observed data for the reservoir elevation shows that the elevation in the reservoir declines during the autumn and winter. The reason for the decline is due to continuing hydropower generation with lower inflows to the reservoir. The participants noted certain conditions that were not included in the model. The actual maximum flood level in the Dniestrovsky reservoir is 122.4 meters. Dam safety issues prevent the elevation from reaching 125 meters. There is an agreement to provide 120-130 cubic meters per second ecological-sanitary flow to Moldova. Minimum releases from the reservoir are a function of reservoir elevation. If reservoir elevation is low, minimum flow is lower. Figure 1: Dniestrovsky reservoir guide curve, Source: Draft Operation Manual for the Dniester Reservoir (Second Edition) (ПРОЕКТ «ПРАВИЛ ЭКСПЛУАТАЦИИ ДНЕСТРОВСКОГО ВОДОХРАНИЛИЩА» (ВТОРАЯ РЕДАКЦИЯ)). 5 Real Time Operations The participants acknowledged that automation of real-time data is necessary for forecasting reservoir inflows. The group discussed the current status of hydro-meteorological data. Manual monitoring is done throughout the basin in 8-20 hour periods. Data are then submitted to a processing center and stored in a database. There are three working automation water level monitoring stations on the Dniester in Ukraine. Twenty automation monitoring stations are installed in the Dniester basin in Moldova; however, their integration into one system is required for further use. Hanbali discussed with participants how to develop scripts or utilities to automate data entry and participants confirmed that there is capacity within the present agencies and institutions to develop these scripts or utilities. However, this need does not seem to be understood at the decision making levels and, therefore, obtaining funding for this effort is a significant challenge. The Ukrainian Centre of Environmental and Water Projects shared their currently ongoing evaluation of software programs for models integration. One of the options they are evaluating, the Community Hydrologic Performance System (CHPS), already incorporates ResSim. They will also now consider the HEC-RTS suite presented by Hanbali. Planning There was misunderstanding relating to the terminology regarding reservoir operating constraints. The USACE team used the term ‘rules’ to refer to any operating constraints tested within the model. The Ukrainian participants pointed out that constraints can only be considered rules if they have been approved. The USACE team attempted to correct their terminology for the remainder of the workshop and in this report. The need to re-evaluate the current operating constraints for the Dniestrovsky reservoir was debated. Current reservoir regulation follows the operation manual from 1987, which had been approved by the relevant authorities. These constraints are not static, but they have been updated for new conditions such as environmental needs and new hydropower plants. There is an effort underway to revise the operation manual for reservoir operations. Many participants noted that this type of tool was not available in the past, and it would be a good opportunity to develop reservoir constraints based on more detailed analysis using the reservoir simulation model. The need to assess and incorporate environmental needs into operations was discussed in great length. The reservoir managers currently conduct artificial spring floods to support spawning, but some ecologists thought these flows were not sufficient. One ecologist said an appropriate flow was 500 cubic meters per second in the delta for at least one month during the April and May period. The Dniestrovsky reservoir has a constraint on the rate of change of its elevation to 10 centimeters per day. This constraint can sometimes limit the amount of the spring artificial flood; however, during several last years the elevation change was equal to 15-19 centimeter or even higher per day due to the needs of artificial spring floods. In addition to fisheries, there are two downstream Ramsar wetlands in the delta that must be considered. It is needed that the model pays special attention to incorporating ecological demands. In Odessa, there are 600 tons of fish caught at the estuary and some agriculture depends on flows from the Dniester. The model could be used to show the effects of reservoir operations on these economic indicators. The ResSim model could be used to evaluate climate change adaptation. Climate change may bring more dry years in addition to larger floods. The ResSim model should be used to evaluate 6 how the system can be managed during drought. Reservoir management can be a tool for climate adaptation. Discussions following the Ukrainian Hydrometeorological Centre presentation provided suggested planning uses for the ResSim model. For short term planning, the model can be used to prepare operations based on monthly forecasts provided by the Centre. The Centre also does short term forecasting for floods with a lead time of 2 days. For long-term planning, the model can be used to develop statistics to evaluate water supply, hydropower, environmental, and flood risk management options. Expectations and Future Applications of the Model The participants provided the following expectations and future applications of the model: Provide a quantitative approach to solve challenges. Key challenges include climate change and increased demands. Scenario analysis can be conducted to develop optimal operating decisions. Scenario time frames can range from seasonal to annual forecasts Improve communications across sectors for water management in the transboundary basin. Evaluate existing and proposed operating constraints for the Dniestrovsky Reservoir. The operation manual is currently being revised with a deadline of September 2015, so the model is very timely. It is important to use the model as a tool for improvement of the reservoir operation manual as well as for wise management of water resources in the basin (incl. flood mitigation). It was mentioned that there is a need for wide discussion of the operation manual with relevant stakeholders. Improve operations for flood risk management. The model can be used to provide a more operational emergency response to flood forecasts in real-time as well as to conduct a planning assessment to evaluate alternative flood storage levels. Bring awareness to the effects of reservoir operations on environmental requirements. This program might be an incentive for cooperation for regulating upstream and downstream users, with fish being an important downstream user. It is obvious that the model is dealing with complex inter-sectoral issues where different aims should be analyzed and prioritized; however, in any case environmental component should not be neglected. The participants identified the following future needs: Economic indicators to evaluate the effects demonstrated by the model. The indicators could include fisheries and agriculture. Inclusion of bank and levee enhancements within the model. Improvement of hydropower component to account for time scale lower than daily. Continuation of the inter-sectoral dialogue on management of the reservoirs and analysis and assessment of priorities. Training of a designated working group within the basin for the ResSim model and further involvement of local experts into modeling. Moldova Feedback on Dniester ResSim Model The participants made comments on the model and recommended some changes. There were several comments about the maximum flood level of 125.0 meters in Figure 1. During the 2008 flood, the Moldovans wanted the water in the Dniestrovsky reservoir to be raised to the elevation of 125.0 meters, but the reservoir managers did not exceed 123.0 meters. The managers said the reason was fear of a breech. 7 Revisions to the water withdrawal locations were provided. The water supply for Chisinau is at Vadul lui Voda downstream of Dubasari. Downstream irrigation diversions are at Grigoripol, Slobozia and Stefan Voda. The number of irrigated hectares in Moldova (including Transnistria) has decreased from around 316,000 hectares in 1990 to about 10,000 hectares in 2000. The average number of hectares to be irrigated in 2015 is about 20,000. The following table provides a percentage of irrigation needs per month. In the long-term perspective (about 50 years) total number of hectares to be irrigated in both Moldova and Ukraine could be 275,000 ha. Table: Percentage of annual irrigation withdrawal occurring in each month. April May June July August Sept Irrigation Withdrawals 5% 15 % 15 % 25 % 25 % 15 % The operation manual says the minimum flow below Dubasari is 100 cubic meters per second. In the 1980s, the minimum sanitary flow was 150 cubic meters per second. Expectations and Future Applications of the Model The participants discussed expectations and future applications of the model: The application for both planning and real-time operations for flood risk management was discussed in detail. The models can be used to provide quantitative assessments for decision makers. The model can be used to facilitate communication to bridge the gap between scientists, engineers, and decision makers. Dubasari Reservoir is essentially a transit reservoir. Therefore, the significant operations primarily exist only at the Dniestrovsky reservoir. The Ukrainian Interdepartmental Commission that sets monthly releases must be involved in discussions regarding the use of the reservoir model. In Moldova, there is also an emergency commission that coordinates operations during floods. One representative commented that the two countries are already working to collaborate in flood mitigation and the ResSim model provides an additional opportunity and additional data for decision making. The Ukrainian Interdepartmental Commission decides the amount of reservoir releases to support environmental needs. The releases for environmental needs (including spring artificial flood) can be modeled using the ResSim model. The model can be used to conduct climate change analyses. Extreme events are expected to occur more frequently in our future, putting sectors at risk in the Dniester River Basin; therefore, climate change scenarios should be taken into account. Researchers in Moldova are currently using HEC-RAS models to assess flood risk in Moldova. The ResSim model can be linked to the hydraulic modeling being done within the basin to provide an integrated assessment of flood impacts of reservoir operations. Participants expressed some concerns regarding model implementation. Without the necessary political and legal instruments, it is difficult to sustainably implement such a model. There needs to be political will to improve and realize these types of models. Furthermore there is no institution in Moldova with sufficient skilled labor or funding to implement the model. The following future needs were identified by participants: Forecasting or runoff from precipitation is necessary to support real-time operations. There is a need for automated data collection systems. A communication network is required to support automation of data. 8 There is a need for prioritization of different basin sectoral objectives which depend on the reservoir management and their economic assessment. Compromise could be hard to achieve (e.g. water supply and flood mitigation versus hydroenergy); however, the model could be a tool for that. Identification and training of a working group to support model development and implementation. There was much discussion regarding who will be the user of the model. One suggestion was that researchers will study and develop it while agencies use the final output to make decisions. However, Moldova does not have a joint group regarding this matter (e.g. there is no Institute of Hydrology). Additionally, on the technical side, IT specialists and hydrologists are needed. The Institute of Hydrology could be an option; however, funding is a challenge for this. It was proposed that training be conducted at universities, since it is an open source software program. The universities can continue to pass the information along to new students. Other participants prefer that people of all ages be trained. In any case, establishment of the local team (at least mobile) will be very beneficial for further development and use of the model. More decision makers should attend future workshops to bridge the gap between scientists and policy makers. Inter-sectoral dialogue on both national and transboundary levels is necessary. Hourly data from the power plant are needed for real-time operations. There is a need for additional research including taking cross-sections at the Dniester and Dubasari reservoirs. At the workshop in Ukraine it was mentioned that such activities were implemented in summer 2014 in Ukraine. It is also necessary to have better information about the timing and magnitude of peak flows. It would be good to develop a program for smaller dams for other rivers in Moldova, which are failing or filling with silt. Key Outcomes The following key outcomes resulted from the workshop: 1. Working groups in Ukraine and Moldova reviewed the Dniester River reservoir simulation model, its data and assumptions, and provided recommendations for its future use. The IWR/AGWA/ICIWaRM team vetted the data and assumptions used to develop the multi-reservoir model with subject matter experts from both countries. In addition to vetting the model, the team provided hands-on experience with the ResSim program and Dniester River Basin model to ensure that the local working groups understood the capabilities of the model for future use. 2. Detailed discussions of both planning and real-time applications of the model were conducted by both the Moldovan and Ukrainian working groups. In regards to real-time applications, the Ukrainian Hydrometeorological Centre is interested in developing a real-time forecasting system for the Upper Dniester River Basin, where the majority of precipitation occurs. The IWR/AGWA/ICIWaRM team was able to provide suggestions based on the USACE existing software programs and examples of how to incorporate the ResSim model developed under this work. Additionally, both the Moldovans and Ukrainians participated in model demonstrations for real-time decision making under flood conditions, using the 2008 flood as an example. In regards to planning, both countries discussed the application of the model to assess reservoir operations. 9 Future Work 1. Revise the ResSim model based on feedback from the meeting participants. 2. Complete the climate risk assessment. This effort will evaluate alternative climate adaptation strategies. Specific areas of attention include: 1) Evaluation of alternative normal operating levels for Dniestrovsky reservoirs and the tradeoffs between potential downstream flooding with hydropower and water supply objectives; 2) Evaluation of alternative spring artificial floods and tradeoffs with hydropower and water supply objectives; 3) Evaluation of drought contingency operating plans. Future scenarios will include changes in climate and irrigation demand. The goal is to complete the assessment by 31 March 2015 if funding permits. The following work will require additional funding. Scopes of work will be developed for the following activities. 3. Develop an improved model of hydropower operations at a one-hour time step, including operating the Dniestrovsky, Pumped Storage and Buffer power plants as a system. The hydropower model will require new operating requirements for the Buffer Dam to replicate intraday reservoir releases and elevation changes. 4. Develop a real-time reservoir management system. The components of a real-time system include real-time automated data collection systems for precipitation and streamflow, a rainfall-runoff model to provide inflow to the reservoir model, the ResSim reservoir management model, and downstream hydraulic models for flood inundation mapping and flood damage models. Investment in an automated data collection system would be needed. This work would be done in collaboration with the hydrometeorlogical agencies in both countries to develop hydrological forecasting models. In addition, the National Academies of Science in both countries have expertise in hydrologic, hydraulic, and flood inundation models that can support this work. 5. Continue building capacity (including training for relevant specialists) to use HEC ResSim and other software to support 1) use of the ResSim model for short-term and long-range planning, and 2) development and implementation of a real-time reservoir management system. The following partners for collaboration were identified for future work: Ukrainian Hydrometeorological Centre and State Hydrometeorological Service of Moldova– forecasting and potential development of hydrologic models National Academy of Sciences: Ukrainian Centre of Environmental and Water Projects and Institute of Geography and Ecology in Moldova – hydrologic, hydraulic and flood inundation models and model integration. Hydropower Company NEK “UkrEnergo” – hydropower operations and reservoir operations manual. 10