Se ha denunciado esta presentación.
Utilizamos tu perfil de LinkedIn y tus datos de actividad para personalizar los anuncios y mostrarte publicidad más relevante. Puedes cambiar tus preferencias de publicidad en cualquier momento.

Comparison of timeslicing approaches: a case study using UK TIMES

47 visualizaciones

Publicado el

Comparison of timeslicing approaches: a case study using UK TIMES

Publicado en: Medio ambiente
  • DOWNLOAD FULL BOOKS, INTO AVAILABLE FORMAT ......................................................................................................................... ......................................................................................................................... 1.DOWNLOAD FULL. PDF EBOOK here { https://tinyurl.com/yxufevpm } ......................................................................................................................... 1.DOWNLOAD FULL. EPUB Ebook here { https://tinyurl.com/yxufevpm } ......................................................................................................................... 1.DOWNLOAD FULL. doc Ebook here { https://tinyurl.com/yxufevpm } ......................................................................................................................... 1.DOWNLOAD FULL. PDF EBOOK here { https://tinyurl.com/yxufevpm } ......................................................................................................................... 1.DOWNLOAD FULL. EPUB Ebook here { https://tinyurl.com/yxufevpm } ......................................................................................................................... 1.DOWNLOAD FULL. doc Ebook here { https://tinyurl.com/yxufevpm } ......................................................................................................................... ......................................................................................................................... ......................................................................................................................... .............. Browse by Genre Available eBooks ......................................................................................................................... Art, Biography, Business, Chick Lit, Children's, Christian, Classics, Comics, Contemporary, Cookbooks, Crime, Ebooks, Fantasy, Fiction, Graphic Novels, Historical Fiction, History, Horror, Humor And Comedy, Manga, Memoir, Music, Mystery, Non Fiction, Paranormal, Philosophy, Poetry, Psychology, Religion, Romance, Science, Science Fiction, Self Help, Suspense, Spirituality, Sports, Thriller, Travel, Young Adult,
       Responder 
    ¿Estás seguro?    No
    Tu mensaje aparecerá aquí
  • Sé el primero en recomendar esto

Comparison of timeslicing approaches: a case study using UK TIMES

  1. 1. Comparison of timeslicing approaches: a case study using UK TIMES Paul Dodds 7 June 2019, ETSAP Workshop, Paris, France UCL Energy Institute
  2. 2. Introduction 1. Hourly (electricity) load curves for the future, for using in TIMES and other models. 2. Impacts of adopting different approaches to timeslicing – 1, 6, 16, 192 timeslices. Case study using the UK TIMES model.
  3. 3. Hourly (electricity) load curves for the future Long-term decarbonisation costs and strategies are sensitive to the shape of future load curves, because peakier load curves require higher capital investments in flexible technologies. 1. Future load curve projections for dispatch models have been produced by adding on heat and transport demands to existing load curves. Yet energy system models suggest a range of system loads, and hence the overall load curve, could change (e.g. LED lights). 0 10 20 30 40 50 60 Electricityload(GW) 2015 2030 2050 2. Energy system models have low temporal resolutions and tend to flatten the future load curve in order to reduce costs. It is unlikely that that this is possible in reality. 3. It is not clear that either approach identifies load curves that would occur in extreme weather periods (e.g. 1-in-20 peak winter demand).
  4. 4. Approach in this study Use high-resolution electricity load curves for demands across the economy in 2010 to derive high- resolution load curves for the future, based on energy system outputs from a 192-timeslice model. 1. Decompose 2010 electricity loads for each energy service demand. 2. Use these to set TIMES load curves for each demand (with other data for heating demands, etc.). 3. Examine how these demands might change in the future using an energy system model. 4. Create electricity load curves for 2030 and 2050 based on the 2010 loads and the energy system model scenario projections.
  5. 5. Decomposition of the 2010 load curve • Annual electricity consumption for each sector based on DUKES data. • Load curves derived primarily using Elexon load factors for each month, days of the week, and hours in each day. • Exelon factors were edited to try to improve the fit to 2010 data, when accounting for temperature errors, etc. • Some data also taken from algorithms in the Estimo model. • The actual load curve is estimated by adding decentralised generation to the National Grid national demand.
  6. 6. Actual and decomposed load curve in 2010 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 01 Jan 31 Jan 02 Mar 02 Apr 02 May 02 Jun 02 Jul 01 Aug 01 Sep 01 Oct 01 Nov 01 Dec 31 Dec Load(MW) Actual demand Modelled Demand
  7. 7. Discrepancy between the decomposed and the actual load curve through the year -15,000 -10,000 -5,000 0 5,000 10,000 15,000 20,000 25,000 30,000 01 Jan 31 Jan 02 Mar 02 Apr 02 May 02 Jun 02 Jul 01 Aug 01 Sep 01 Oct 01 Nov 01 Dec 31 Dec Differencebetweendecomposedandactualload(MW)
  8. 8. Discrepancy between the decomposed and the actual load curve by time of day -15,000 -10,000 -5,000 0 5,000 10,000 15,000 20,000 25,000 30,000 Differencebetweendecomposedand actualLoad(MW)
  9. 9. Average daily load in 2010 0 10,000 20,000 30,000 40,000 50,000 60,000 01 Jan 31 Jan 02 Mar 02 Apr 02 May 02 Jun 02 Jul 01 Aug 01 Sep 01 Oct 01 Nov 01 Dec 31 Dec Load(MW) Actual Modelled
  10. 10. Discrepancy in 2010 as a function of daily air temperature -15000 -10000 -5000 0 5000 10000 15000 20000 25000 30000 -10 -5 0 5 10 15 20 25 30 Discrepancy(decomposed-actualload)(MW) Average daily temperature (degC)
  11. 11. 0% 10% 20% 30% 40% 50% 60% -15,000 -10,000 -5,000 0 5,000 10,000 15,000 20,000 25,000 30,000 Fractionoftotalelectricitydemandinsector Electricity demand discrepancy (model - actual) (MW) % Res % Ser % Oth Linear (% Res) Linear (% Ser) Linear (% Oth)
  12. 12. Decomposition of the load curve in 2010 – average daily load 0 10,000 20,000 30,000 40,000 50,000 60,000 01 Jan 01 Feb 01 Mar 01 Apr 01 May 01 Jun 01 Jul 01 Aug 01 Sep 01 Oct 01 Nov 01 Dec UKload(MW) Residential weather Residential non-weather Service weather Service non-weather Transport Agriculture Industry Oil refineries Hydrogen
  13. 13. Decomposition of the load curve in 2010, for an 80% GHG reduction but no other constraints – peak daily load 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 01 Jan 01 Feb 01 Mar 01 Apr 01 May 01 Jun 01 Jul 01 Aug 01 Sep 01 Oct 01 Nov 01 Dec UKload(MW) Residential weather Residential non-weather Service weather Service non-weather Transport Agriculture Industry Oil refineries Hydrogen
  14. 14. Future load projections in UK TIMES STEP 2: • The decomposed load curves for each demand are used to calibrate UK TIMES energy service demands. • A scenario of the future is produced using UK TIMES. STEP 3: • The changes in each demand are output to an Excel workbook. • The decomposed load curves for 2010 are multiplied by the future demands to estimate the future load curve. • Micro-CHP and industrial CHP generation is assumed to be heat-led. Other CHP is assumed flat (i.e. with boilers or other heat-only devices, and heat storage, coping with peaks. • Some EV demand is assumed flexible. An algorithm first allocates this to overnight low periods, then any low period, then to flatten the load curve.
  15. 15. Decomposition of the load curve in 2050, for an 80% GHG reduction but no other constraints – average daily load 0 20,000 40,000 60,000 80,000 100,000 120,000 01 Jan 01 Feb 01 Mar 01 Apr 01 May 01 Jun 01 Jul 01 Aug 01 Sep 01 Oct 01 Nov 01 Dec UKload(MW) Residential weather Residential non-weather Service weather Service non-weather Transport Agriculture Industry Oil refineries Hydrogen
  16. 16. Decomposition of the load curve in 2050 – peak daily load 0 20,000 40,000 60,000 80,000 100,000 120,000 01 Jan 01 Feb 01 Mar 01 Apr 01 May 01 Jun 01 Jul 01 Aug 01 Sep 01 Oct 01 Nov 01 Dec UKload(MW) Residential weather Residential non-weather Service weather Service non-weather Transport Agriculture Industry Oil refineries Hydrogen
  17. 17. 2. IMPACTS OF ADOPTING DIFFERENT APPROACHES TO TIMESLICING
  18. 18. 0 10 20 30 40 50 60 00:30 01:30 02:30 03:30 04:30 05:30 06:30 07:30 08:30 09:30 10:30 11:30 12:30 13:30 14:30 15:30 16:30 17:30 18:30 19:30 20:30 21:30 22:30 23:30 Demand(GW) Time Ending National Grid (NETS) Figure 2.1 - Summer and Winter Daily Demand Profiles in 2010/11 Winter Maximum Typical Winter Typical Summer Summer Minimum Evening peak Timeslicing approaches • UK MARKAL used 6 timeslices (3 season, 2 daynite) • UK TIMES uses 16 timeslices (4 season, 4 daynite) Night Day Late evening • What timeslicing approach finds the sweet spot between resolving temporal variability and minimising data requirements and running time? • 1 timeslice (20 seconds) • 6 timeslices (2 minutes) • 16 timeslices (8 minutes) • 192 timeslices (4 hours, workstation only)
  19. 19. Modelling approach • Base model is annual only. • Timeslicing data is introduced only by SysSettings and by scenario files. • Hourly energy demands are aggregated to produce consistent data for each timeslicing approach. • Different electricity reserve capacity fractions are calculated for timeslicing approach. • Approach partly made possible by the release of TIMES Version 4.3.2 (March 2019) – which added support for varying timeslice cycles. Previously, processes defined at timeslice levels that were not in SysSettings were converted to ANNUAL processes. Now, they move to the next most aggregated timeslice level. • For UK TIMES, WEEKLY  SEASON.
  20. 20. 192 timeslices 0% 20% 40% 60% 80% 100% 1 3 5 7 9 11 13 PV: Winter Spring Summer Autumn 0% 20% 40% 60% 80% 100% 1 2 3 4 5 6 7 8 9 10 11 12 Offshore wind 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 0% 20% 40% 60% 80% 100% Onshore wind • Designed to represent variable renewable generation • 4 typical days per season with varying • 12 periods per day
  21. 21. Daily load curve variability by season in 2010 • In each season, each representative day has the same demands. • These change between seasons. • Hence the model ignores demand variability within seasons 0 10 20 30 40 50 60 70 Spring 0 10 20 30 40 50 60 70 Winter 0 10 20 30 40 50 60 70 Summer 0 10 20 30 40 50 60 70 Autumn Load(MW)Load(MW)
  22. 22. Objective function differences between the four versions Timeslices ObjZ Difference 1 6,942,001 6 6,903,674 -0.6% 16 6,872,681 -1.0% 192 6,891,317 -0.7% Lower because a high winter demand stress period is not represented?
  23. 23. CO2 emissions by sector in 2050 -100 -50 0 50 100 150 200 1 TS 6 TS 16 TS 192 TS MtCO2e Non-energy use Upstream Processing Hydrogen Transport Residential Industry Electricity Services Agriculture
  24. 24. Electricity generation in 2050 0 500 1,000 1,500 2,000 1 TS 6 TS 16 TS 192 TS PJ/year Year
  25. 25. Electricity generation in 2050 0 500 1,000 1,500 2,000 1 TS 6 TS 16 TS 192 TS PJ/year Year Storage output Net Imports Fuel cells Hydrogen Nuclear Hydro Geothermal Wave Wind Tidal Solar Biomass CCS
  26. 26. Residential heat in 2050 0 200 400 600 800 1,000 1,200 1,400 1 TS 6 TS 16 TS 192 TS PJ/year District heat Standalone water heater Standalone air heater Night storage heater Fuel cell micro-CHP Stirling engine micro-CHP Solar heating and heat pump Solar heating and boiler Hybrid heat pumps Heat pumps Hydrogen boiler Biomass boiler Electric boiler Coal boiler Oil boiler Gas boiler
  27. 27. Road transport end-use technologies in 2050 0 100 200 300 400 500 600 700 800 900 1 TS 6 TS 16 TS 192 TS Bvkm Van Petrol HGV Hybrid Hydrogen HGV EV Car Petrol Car EV
  28. 28. Comparison of peak daily load curves in 2050 0 20,000 40,000 60,000 80,000 100,000 120,000 01 Jan 01 Feb 01 Mar 01 Apr 01 May 01 Jun 01 Jul 01 Aug 01 Sep 01 Oct 01 Nov 01 Dec Peakdailyelectricitydemand(MW) TS1 TS6 TS16 TS192
  29. 29. Conclusions • Finding credible load data is challenging. • The approach in this study is not ideal, but produces reasonable load curves for higher-resolution dispatch models. It’s straightforward to explain and the error can be gauged. • Using higher temporal resolutions does not greatly change the model outputs for this scenario. • Even with 192 timeslices that are optimised for different renewable output days, the model does identifies a much lower need for energy storage than in some hourly studies. • The next step is to repeat the storage analysis, with the same input data, in the hourly highRES model. 1 TS 6 TS 16 TS 192 TS 1.37 1.37 2.26 1.38 Electricity storage capacity (GW)

×