Task 1.9 (5B): Develop a Model of Photovoltaic Energy Systems for IRW Block

Transcription

1 Task 1.9 (5B): Develop a Model of Photovoltaic Energy Systems for IRW Block PI: Prof. Dionysios Aliprantis RA: Chengrui Cai UC Davis, CA May 13, 213 D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

2 Overview Task description: Develop a model for a photovoltaic (PV) energy system, either of the residential rooftop or the large utility-size type. This PV system will be represented using an appropriate physics-based model, e.g., a stochastic model that captures its physical characteristics appropriately. Milestone: The (sampled) distributions of stochastic process model match historical data within 1% on the first two distributional moments. Task progress: 1% complete. Notes: 1. Questions raised in the last meeting 2. Use of historical irradiance data D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

3 Overview What do we want to achieve? 5 scenarios of aggregated hourly PV power (MW) for each ISO-NE load zone (bus) on each day in 211 Hourly solar irradiance time series for all PV systems Locations and rated capacities of all PV systems Irradiance-to-electric power conversion model PV power scenarios consistent with the wind power and load scenarios From the same weather model With same scenario probability D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

4 Overview Data and technical approach What resources are we using? 3TIER provided weather scenarios for 211 SolarAnywhere irradiance database 1 (freely available) NREL OpenPV database 2 (freely available) : Irradiance scenario development PV siting information Irradiance-to-electric power conversion model 1 solaranywhere.com 2 openpv.nrel.gov D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

5 Irradiance scenario development Irradiance scenario development D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

6 Irradiance scenario development Weather scenarios provided by 3TIER Locations of 8 airports 5 scenarios for each day Meteorological variables: global horizontal irradiance ground level wind speed dew-point, temperature clearness index, pressure ID Location Latitude Longitude KBDL Hartford, CT N W KBDR Bridgeport, CT N W KBOS Boston, MA N W KBTV Burlington, VT N W KCON Concord, NH N W KORH Worcester, MA N W KPVD Providence, RI N W KPWM Portland, ME N W D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

7 Irradiance scenario development Irradiance scenarios and probabilities GHI (W/m 2 ) Sc #1 Sc #2 Sc #3 Sc #4 Sc #5 Sc #6 Sc #7 Sc #8 Sc #9 Sc #1 Probability Hour (a) Scenario # (b) An example of the 3TIER solar irradiance scenarios for Hartford, CT for Jan. 2nd, 211. The first 1 scenarios with the highest probability are plotted in (a) with varying line thickness, according to the probability of occurrence shown in (b). D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

8 Irradiance scenario development SolarAnywhere irradiance database Hourly irradiance data: global horizontal irradiance (GHI) direct normal irradiance (DNI) diffuse horizontal irradiance (DHI) Entire U.S. from 1998 to 29, i.e., 4383 days, with a spatial resolution of.1 (around 1 km) for both latitude and longitude D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

9 Irradiance scenario development Match the SolarAnywhere with 3TIER data To identify similar days from the SolarAnywhere database such that the solar irradiance (specifically the global horizontal irradiance) pattern over these 8 locations is close to the 3TIER scenarios. Algorithm 1 The solar scenario generation algorithm for a particular day D 1: input: 2: T: matrix 3: S: matrix 4: output: 5: I: 5 1 vector 6: variable: 7: E: matrix 8: for j 1 to 5 do 9: for k 1 to 4383 do 1: E j,k = 8 i=1 ( T:,i,j S :,idx(i),k 2 / T :,i,j 2 ) /8 11: end for 12: I j = argmin E j,k k [1,4383] 13: end for note: 1. T :,i,j is the vector of 24 hours of GHI data of i-th location for j-th scenario in the 3TIER data. 2. symbol : in the subscript means the entire data on that dimension of the matrix. 3. idx(i) is the index of the i-th location in the SolarAnywhere data. D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

10 24 rows (hours) 3TIER 5 pages (sc.) T :,i,j 4383 pages (days) Irradiance scenario development SolarAnywhere 8 columns (locations) Averaged error: 8i=1 T :,i,j S :,idx(i),k 2 T :,i,j 2 E j,k = 1 8 Selected day: I j = argmin k [1,4383] E j,k 24 rows (hours) S :,idx(i),k 2837 columns (locations) D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

11 Solar irradiance scenario generation GHI (W/m 2 ) Clear day scenario Cloudy day scenario KBDL KBDR KBOS KBTV KCON KORH KPVD KPWM GHI (W/m 2 ) Red: SolarAnywhere Blue: 3TIER KBDL KBDR KBOS KBTV KCON KORH KPVD KPWM D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

12 PV siting information PV siting information D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

13 PV siting information NREL OpenPV database of PV system info in US Capital cost Rated capacity Location (latitude and longitude) D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

14 Current situation PV siting information MWdc ME, NH and RI VT CT MA 5.65 Historical trend of installed PV capacity from 27 to in ISO-NE to 211 data from IREC (Interstate Renewable Energy Council) Solar Market Trends Reports 212 data from OpenPV database Year D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

15 PV siting information Process OpenPV data for SolarAnywhere 46 N 46 N ME NY VT NH 44 N 44 N MA 42 N 42 N CT 72 W 72 W RI 7 W 68 W Each blue dot is a PV system tiles (1 km 1 km) D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting W 7 W Density of PV capacity: kw/km May 13, / 22

16 PV siting information Magnified MA, CT and RI area 42.5 N 42. N NY MA 41.5 N 41. N 73.5 W 72.5 W CT RI 71.5 W 7.5 W Density of PV capacity: kw/km 2 D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

17 Irradiance-to-electric power conversion model Irradiance-to-electric power conversion model D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

18 Irradiance-to-electric power conversion model From irradiance to electric power Latitude (degrees) ISO-NE 8 load zones Source: Longitude (degrees) Each point is the center of a tile. D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

19 Irradiance-to-electric power conversion model Power conversion model P = C η G G n P : hourly average electric power in MW AC generated by PV systems in a tile C : total PV capacity in MW DC in a tile (from OpenPV) η : PV derating percentage 3 (e.g., wiring losses, shading, aging, dirt, etc) G : hourly average solar irradiance incident on the PV arrays in W/m 2 (function of DNI and DHI from SolarAnywhere) G n : STC (standard test condition) constant of 1 W/m 2 3 default value is.77, as suggested by NREL PVWatts model. D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

20 Irradiance-to-electric power conversion model Aggregated PV power for each load zone A winter clear day A summer cloudy day MWAC Maine Vermont NE Mass Connecticut New Hampshire WC Mass SE Mass Rhode Island MWAC Maine Vermont NE Mass Connecticut New Hampshire WC Mass SE Mass Rhode Island D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

21 Summary Conclusion Develop scenarios of photovoltaic power in ISO-NE region for the stochastic unit commitment problem. 5 scenarios for each of the 8 load zones (buses) 365 days in 211 Correlated with wind and load profiles The technical approach includes three components. Irradiance scenario development (using historical measurements) PV siting information (based on actual PV system locations) Irradiance-to-electric power conversion model (physical-based model) D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22

22 Q & A Thank You Questions? Contacts: Prof. Dionysios Aliprantis (dali@iastate.edu) Chengrui Cai (ccai@iastate.edu) D. C. Aliprantis, C. Cai (ECpE, ISU) ARPA-E Project Meeting May 13, / 22