Improving the model of training view from the outside

Transcription

1 Improving the model of training view from the outside Fedor Veselov, Energy Research Institute of RAS Workshop to Exchange Experience among Trainers on the IAEA s Models for Energy System Planning Vienna, July 2016

2 A. Recent training process how to maximize the long-term effect from new knowledge B. Sharing the experience from decision making process C. Some experiments with self-training 2

3 Recent results of training activity view from outside Quantitative results are impressive: - MESSAGE is distributed in near 40 countries - Hundreds of people know about the model and have modeling skills (at different level) - IAEA provides unique opportunities of free access to the model supplemented by well-developed the system of learning and training But there is high risk at the country level that modeling skills will be dissappeared with a time To eliminate these risks, it is extremely important to integrate the model into the regular system of energy planning at the national/regional level. Training courses may partially help to improve the attractiveness of the model to decision makers Trained people may stop to use MESSAGE and share the knowledge if they will lost scientific or/and financial motivation How many people/teams use the MESSAGE on a regular basis as an investigation tool? Can this longterm effect of training be increased? How the situation can be changed to make the model more attractive as a tool for the decision making process? 3

4 How to improve the competitiveness and credibility of the MESSAGE as a future investigation tool in the member states? What can be done by PESS office (taking into account labor and financial resources) Enhance capabilities of the model in new versions (e.g. cloud calculations) Improve the input-output interface (like for WASP and MAED) Finalize the model manual Provide the more (and more open) access to the training materials and simple cases Make the assistance tool (TSES) more effective in terms of faster reply. What can be done during the training activity - More attention to the enhanced trainings that allow to obtain and apply new skills to the real energy system planning tasks - Focusing on the professional background and position of the trainees - Speak more about the general energy planning issues to show the real role of the modeling - Teach to present not the optimization results only but the model logic, rationale for the obtained optimal solutions as well as boundary and risks issues - Extended application of the model to model other production chains outside of energy sector (A.Galinis 2015) 4

5 How to improve the competitiveness and credibility of the MESSAGE as a future investigation tool in the member states? - Focus on the background and position of trainees - -Position Trainees of are trainees: already working in the area of energy sector (or Speak -separate people not industry) only from about academic/corporate and the they model must operation have institutes the issues, basic and knowledge but also in about energy universities its sector logic as structure/statistics/balance can well form as and the maintain nature of the energy at permanent the level planning of groups for separate using the industry model or and sector continue as a whole the process of sharing the - Ideally knowledge. the energy But it forecasting/strategic is important to help planning/investment them to be successful at planning the Model highly must operator competitive be the skills of (start/run/cases/input the market core professional energy /output/errors) forecasts activity of the trainees - -But people if the Model trainees from data public preparation are from authorities different and input (ministries, skills areas (power commissions, sector, statistic etc.) or service, utilities regulator can improve office, personal fuel industries, professional environmental organizations, background, Reference etc.) but often data maybe and they information it are is better not resources interested try to compare in maintaining the integrated the regular modeling team application: in advance (maybe as a special requirement - they may during change the organization the position of in training 2-3 years course) Examples to improve experience - The -alternative their functionality is to prepare may the not trainings include regular with special modeling focuses (to the exercises power sector, fuel industry, RES or environmental issues) Model mathematics, model and matrix structure - important exception (not in all countries) some - Effect of special teaching groups/departments the heterogeneous in groups the structure of people of public is not so Different technologies of system studies (multi-case analysis, much in authorities the sensitivity long-term analysis, responsible because risk assessment, for the providing new marginal knowledge the energy cost analysis, about forecasting the etc.) model background, often is not comply without with the their assistance professional external areas What experts/consultants/organizations if grounds for decision makers how to make the results credible for them? 5

6 Building the bridge between energy modelers and decision makers In may cases main problem is to overcome the gap between the energy planners and energy authorities/companies. It is critically important for the future maintaining of the acquired modeling capacity and its integration in the real decision making process. And it may be a part of training! Energy planners (operators of the models) Academic institutes Universities Analytical departments Quantified forecasts of energy sector development based on the formal mathematic modeling and approximation of energy sector structure Believe or not believe? Requirements for the arguments for decision making: Alternatives? Why this is the best? Affordability vs optimality Clear effects, risks, externalities 6 Energy decision makers Ministry Commission Public authorities HQ of energy companies

7 A. Recent training process how to maximize the longterm effect from new knowledge B. Sharing the experience from decision making process Simple economic comparison of energy technologies based on the discounted (or levelized) generation costs Multi-scenario analysis of the alternative investment strategies: savings of resources and impact on total costs of energy supply Analysis of price effects based on primal and dual solution C. Some experiments with self-training 7

8 Staged approach to the generation capacity structure forecasting Screening analysis of the typical decisions of existing plant rehabilitation and new construction options (selection of the preferred generating technologies) Modeling tool spreadsheet calculator of discounted costs of construction, operation and decommissioning Criterion per kwh discounted generation costs for each technology (LCOE, EGC) It may be FINPLAN or simple spreadsheet calculator System evaluation of energy balanced and economically efficient variants of power sector development (optimization of scales for preferred generating technilogies) Modeling tool long-term optimization model of power sector integrating all generating technologies and grid expansion (fuel supply as an option) It may be MESSAGE or WASP Criterion minimum of total discounted costs of forecasted balance requirements supply in the planning period (+aftereffects in 15 years) Formulation of the rational generation capacity structure variant Modeling tool simulation models for balance calculations, seasonal and daily operation of power system Criterion least deviations from optimal structure caused by the variants and costs of grid projects, system reliability constraints, unit capacities, predefined (must be built) projects as well as fuel and capital resource limitations 8

9 Input data: cost and price ratios are more important rather than values Relative capex (% to CCGT) Unit size effect on the capex of gas-fired CHP 300% Gas CHP-CC 250% 200% 150% 100% 50% 0% gas CHP- CC % gas CHP GT coal SC plant coal USC plant coal CHP Fuel prices growth to 2035 (% to 2015) Gas Nuclear Coal 2,6 2,4 2,2 2,0 1,8 1,6 1,4 1,2 1,0 0, Unit size, MW Gas/coal price ratio by the key regions of inter-fuel competition Ural South Center

10 Cost-based screening analysis: inter-fuel competition between conventional plant types Changes of electricity generation costs for new generation capacity options (Center integrated power system), 0,01 RUR/kWh CCGT Nuclear Coal SC plant O&M costs fuel costs capital costs BAU (2015) fuel price growth BAU (2015) 20% 20% cost capex of capital reduction reduction BAU (2015) fuel price growth 10

11 Assessment of the alternative nuclear generation development strategies Nuclear development strategies Strategy Content Nuclear capacity to 2035, GW Minimum Substitution of existing units only 31 Average Substitution of existing units and new additions 35.4 limited by the balance conditions Maximum 100% performance of Rosatom roadmap 43.4 Changes in the fuel and capital requirement options (D from Average strategy) Mtce Maximum 499,5 499,4 499,3 499,2 Change in cumulative discounted energy supply costs for different nuclear development strategies, bln USD Maximum Minimum 499,1 499 Minimum Bln RUR Difference in cum.capital requiraments Difference in fuel requirements in ,9 498,8 Average New nuclear capacity in 2035, GW 498,

12 Assessment of the alternative thermal generation rehabilitation strategies (resource/cost saving estimations from modeling) Thermal capacity additions by the rehabilitation scenarios, GW NewTech scenario Mixed scenario PrevTech scenario Increase of new thermal technologies Rehab with same technologies Rehab with new technologies New capacity aditions Changes in the fuel and capital requirement options (delta from Mixed scenario) Mtce Bln USD NewTech PrevTech Difference in cum.capital requiraments Difference in fuel requirements in Change in cumulative discounted energy supply costs for different thermal plants rehabilitation scenarios, bln USD PrevTech Mixed NewTech New thermal technologies, GW

13 Assessment of the alternative thermal generation rehabilitation strategies (prices estimation based on economic analysis of model primal and dual solution) Increasing volumes of new thermal capacities with lower heat rates (and fuel costs) will change the supply curve and tend to lowering spot electricity prices. But this effect must be compared with the change of capacity payments. Spot price, RUR/MWh Decrease of spot prices with the change of supply curve profile as a result of new thermal capacities increase How to estimate price impacts: Spot electricity prices reduced costs of electricity supply constraints (marginal cost of electricity supply) Capacity payments fixed O&M costs and capital charge rates from the model cost function 2,5 2 Wholesale price, RUR/kWh capacity price spot electricity 1,5 1 0,5 Electricity supply, GWh 0 PrevTech Mixed NewTech 13

14 Multy-case optimization of the generating capacity structure Upgrade of existing thermal capacities Development of cogeneration Development of new thermal and nuclear plants Capacity additions by the type of plants, GW Fuel prices Capital costs of upgrade projects Fuel prices Capital costs of new CHP and boilers Fuel prices Capital costs of new nuclear, gas and coal plants Variable factors Electricity demand Investment stimulation for upgrade Heat demand Stimulation of distributed generation for substitution of existing boilers Electricity demand Alternative regional location of greenfield plants Modeling cases Final result Number of modeling cases Investment decisions for the thermal plants, GW Modeling cases Final result Nuclear Gas fired Coal fired CHP New nuclear capacity additions by the regions, GW Modeling cases Final result Modernization and life extension Substitution with higher efficiency New capacity 14 North-West Center Volga South Ural

15 Installed capacity structure (Unified power system) GW ,5 0,0 71,6 High demand case 243,8 250,8 238,9 1,4 3,8 2,6 74,4 71,7 77,6 264,8 5,7 81,6 GW Low demand case 252,9 243,8 243,8 232,5 235,4 1,4 0,0 2,6 3,8 5,7 74,4 71,6 69,4 71,9 74, ,8 85,7 80,9 84,1 89, ,8 85,7 80,9 84,1 89, ,7 50,6 51,7 52,3 54, ,7 50,6 51,7 52,3 52,5 0 26,3 31,7 32,0 33,0 33, ,3 31,7 30,8 31,8 31, Nuclear Hydro+PSP CHP CPP RES Nuclear Hydro+PSP CHP CPP RES Nuclear 11,3% 13,0% 13,4% 13,2% 12,6% Hydro+PSP 20,5% 20,7% 21,6% 20,9% 20,6% CHP 37,3% 35,2% 33,9% 33,5% 33,8% CPP 30,8% 30,5% 30,0% 30,9% 30,8% RES 0,0% 0,6% 1,1% 1,5% 2,2% Nuclear 11,3% 13,0% 13,1% 13,0% 12,2% Hydro+PSP 20,5% 20,7% 22,0% 21,5% 20,8% CHP 37,3% 35,2% 34,4% 34,5% 35,4% CPP 30,8% 30,5% 29,5% 29,5% 29,3% RES 0,0% 0,6% 1,1% 1,6% 2,2% 15

16 A. Recent training process how to maximize the longterm effect from new knowledge B. Sharing the experience from decision making process C. Some experiments with self-training 16

17 Step-by step self-training Irina Tatiana Andrey Recent experience in: - Power sector economics Good (master degree) Good (master degree) Good (master degree) - Power sector balances Good skills, key expert None Good skills, key expert - LP-modeling experience None None Good, creating similar models 17

18 Step-by step self-training Educational resources: Initial e-learning course (5 days) MESSAGE manual (2008 version) Some presentations from IAEA and Internet Single region case Extension to the multiregion case Add load region options Extension to the whole energy sector Plant data input Electricity and heat demand constraint Modeling plant operating modes (also CHP) Modeling capacity decrease/additi on Modeling installed capacity requirements Integration of data from single cases Modeling the transmission lines and electricity exchange Presenting the hourly load data in the best way (aggregation into load regions) Modeling the seasonal/daily variations of supply/demand options Developing the full-scale multiregional energy model Modeling the inter-fuel and technological competition 18

19 Step-by step self-training Direct purpose to enhance the experience of young scientists and obtain (or improve) their modeling skills Indirect purpose to estimate how easily specialists can learn the MESSAGE and use it in their practice starting from zero and without regular external assistance General estimation it is possible, but the way is not very straight and clear: E-learning (english version) is rather comprehensive and useful as a first step But after Manual is detailed but not finished: in some very important parts there are??? Instead of explanations/examples. Parameters are often not explained clear and in formulas sometimes it is necessary to make several iterations to change the value or combination and later understand how the model understand the user choice read the matrix directly to see on row and columns and find the sense of parameters from that side More tested examples are required: it is easier to see and repeat instead of invent a new bicycle again. 19

20 Energy research institute of the Russian Academy of Sciences 20