R e p l a c i n g D i e s e l G e n e r a t i o n w i t h R e n e w a b l e S o u r c e s i n N u n a v u t C o m m u n i t i e s : F e a s i b i l i

Transcription

1 R e p l a c i n g D i e s e l G e n e r a t i o n w i t h R e n e w a b l e S o u r c e s i n N u n a v u t C o m m u n i t i e s : F e a s i b i l i t y S t u d i e s Indrajit Das Claudio Canizares Dept. Electrical & Computer Engineering Waterloo Institute of Sustainable Energy 1

2 Outline Procedure Optimization model Methodology General consideration Study cases Search space Results Main input parameters for each community Optimal solutions for various case studies and observations Conclusions and next steps 2

3 Procedure From utility perspective: 3

4 Optimization Model Objective function: Minimization of the net present value of the sum of capital, fuel, and O&M costs for diesel and RE sources. Constraints: Supply demand balance. Piecewise-linear fuel curves for existing and new diesel generators. Dispatch limit for diesel generators. Piecewise-linear wind power curves. PV generation model based on irradiation and temperature levels. Battery charging and discharging equations and limits. Resource adequacy: Operating reserves. Stand-by operation. Remaining and useful life of existing and new diesel generators, respectively. Solved using GAMS (General Algebraic Modeling System). 4

5 Methodology General considerations: Window of new capacity additions: Diesel: 3 rd to 6 th year. Solar: 2 nd to 5 th year. Wind: 3 rd to 5 th year. Battery: 2 nd to 5 th year. Hourly load profile with 1% annual load growth, solar irradiance, wind speed, and temperature data. Existing and pre-selected new diesel generators, wind, solar, and Liion battery equipment and costs for each community. Stand-by operation of selected existing diesel generators for all communities (currently only in some communities). Operating reserve of 25% and 50% of solar and wind energy generated, respectively, at every hour. Discount rate of 8%, which is somewhat conservative nowadays, especially for government institutions. 5

6 Methodology The work is based on seven case studies as follows: NoRE: Business As Usual (BAU) case involving only diesel generation. S: Only solar energy. W: Only wind energy. SW: Both solar and wind energy. SB: Solar with battery storage. SW: Wind with battery storage. SWB: Both solar and wind with battery storage. Minimization of the Net Present Cost (NPC) for a 10- or 20-year project horizon. 6

7 Methodology Search space: Diesel: All Caterpillar generators of capacities of 320, 520, 800, 1000, 1500, 2000, 3000, 4000, and 6000 kw. Wind: Northern Power System s NPS100 (100 kw), Norvento s ned100 (100 kw), EWT s 250, 500, and 900 kw, and Enercon E70 (2300 kw). Solar: Canadian Solar s panelset of 9.6 kw and First Solar s panelset of 10 kw. Battery: Canadian Solar s Li-ion battery of close to 10 kwh capacity, assuming a unit size of 10 battery packs, i.e. ~100 kwh. Different combinations of the aforementioned search space were used for the various communities. 7

8 Sanikiluaq Hourly profiles daily averages per month for load (for the first year) and wind speed height: 8

9 Sanikiluaq Hourly profiles daily averages per month for solar irradiation and temperature: 9

10 Sanikiluaq Optimal results for 10-year horizon: Observations: Maximum reduction of fuel cost (FC) occurs when wind is deployed. NPCs of optimal total costs for RE cases are less than BAU (NoRE) by ~$2 million (~10 %), thus justifying RE deployment in all cases. Maximum diesel use reduction of about 35 % occurs for the SWB case with a maximum RE penetration of close to 50 %. 10

11 Sanikiluaq Partial optimal results for 20-year horizon: Observations compared to 10-year project horizon: Solar capacity addition remains almost the same, but wind and battery capacity additions increase. Diesel capacity additions increase when no batteries are considered. 11

12 Iqaluit Hourly profiles daily averages per month for load (for the first year) and wind speed height: 12

13 Iqaluit Hourly profiles daily averages per month for solar irradiation and temperature: 13

14 Iqaluit Optimal results for 10-year horizon: Observations: Both solar and battery minimum capacity constraints are required for S and SB cases, thus generating higher NPC of total costs, indicating that these combinations are not cost effective for this community. WB is the most cost effective solution, generating maximum reduction in diesel fuel costs (use) of ~6 % with a 25+ % maximum RE penetration, and total cost savings of $4.6 million (~3.5 %). 14

15 Rankin Inlet Hourly profiles daily averages per month for load (for the first year) and wind speed height: 15

16 Rankin Inlet Hourly profiles daily averages per month for solar irradiation and temperature: 16

17 Rankin Inlet Optimal results for 10-year horizon: Observations: Both solar and battery minimum capacity constraints are required for S and SB cases, thus generating higher NPC of total costs, indicating that these combinations are not cost effective for this community. WB is the most effective solution, generating maximum reduction in diesel fuel use of 35 % at close to 50 % maximum RE penetration, and total cost savings of ~$9 million (~15 %). 17

18 Baker Lake Hourly profiles daily averages per month for load (for the first year) and wind speed height: 18

19 Baker Lake Hourly profiles daily averages per month for solar irradiation and temperature: 19

20 Baker Lake Optimal results for 10-year horizon: Observations: Both solar and battery minimum capacity constraints are required for S and SB cases, thus generating higher NPC of total costs, indicating that these combinations are not cost effective for this community. WB combination reduces the fuel use by 22 % with respect to BAU (NoRE) with a maximum RE penetration of close to 40 %, although it is not cheaper than the W case, yielding total cost savings of $1.3+ million (4.5 %). 20

21 Arviat Hourly profiles daily averages per month for load (for the first year) and wind speed height: 21

22 Arviat Hourly profiles daily averages per month for solar irradiation and temperature: 22

23 Arviat Optimal results for 10-year horizon: Observations: Both solar and battery minimum capacity constraints are required for S and SB cases, thus generating higher NPC of total costs, indicating that these combinations are not cost effective for this community. WB is the best option as it maximizes the reduction in fuel use by ~ 40 %, at a total cost savings of ~ $2.5 million (10+ %), and presents the maximum RE penetration of all communities at close to 60 %. 23

24 Conclusions and Next Steps In all communities, RE deployment reduces fuel consumption. For the best options: Maximum yearly RE penetration from 22 to close to 60 %, which is higher than the pre-feasibility results obtained with HOMER due to approximations in diesel dispatch calculations. Total cost savings from 3.5 to 15 %. Fuel savings from 6 to 40 %. Wind is the preferable RE option for all communities. Battery addition reduces fuel use, but it is an overall more expensive solution than without it for some communities. Of all communities, Sanikiluaq is the only community where SWB is the best option. Final results for a 10-year project horizon were presented with reduced search spaces for all communities to meet workshop deadlines. In the feasibility report, results for a 20-year project horizon and a broader search space for each community will be presented. 24

25 Arviat Partial optimal results for 20-year horizon: 25

26 Rankin Inlet Partial optimal results for 20-year horizon: 26