The URL can be used to link to this page

Home My WebLink AboutNaknek Battery Energy Storage System Scenario Analysis - Dec 2022 - REF Grant 7015022Electric Power Systems, Inc. 3305 Arctic Blvd Suite 201 ■ Anchorage, Alaska 99503 ■ Tel: (907) 522-1953 December 1, 2022 Subject: NEA BESS Scenario Analysis ANALYSIS OVERVIEW The objective of this analysis is to quantify the potential benefits of using a Battery Energy Storage System (BESS) to supplant the use of reserve spin. The specific quantities to be developed by this analysis is the reduction in genset run -hour, reduction in fuel usage, and change in heat recovery capability between two utility operating scenarios: the existing scenario where gensets are staged on with one full genset of reserve spin, and a future scenario where there is no reserve spin (due to the use of a future BESS). ANALYSIS INPUTS The inputs to this analysis are: • CAT genset performance data • Approximately 560 days of electronically logged power delivered to distribution. The genset performance data for a CAT 3512B and a 3516 genset were used: GENSET % LOAD EKW FUEL FUEL JW REJ JW RFI GPH GPHJKW BTU/MIN MBH 3512E 100% 1,015 69,5 0A69 28,321 1,699 3512E 909£ 914 62.2 0.068 25,376 1,553 3512E 80% 312 55.3 0.068 23,544 1,413 3512E 75°% 761 52 0A66 22,464 1,348 3512E 70% 710 48.8 0,069 21,326 1,230 3512E 60r% 609 42,8 0.070 19,165 1,150 3512E 50% 508 37 0.073 17,061 14024 3512E 40% 406 31.3 0.077 14,957 997 3512E 30% 305 25.5 0.084 12,739 764 3512E 25YD 254 22,6 0.039 11,545 693 3512E 201A 203 19.7 0,097 10,293 618 3512E 10% 102 13.4 0.132 7,564 454 3516 100°% 1,135 75,53 0A67 34,691 2,081 3515 913% 1,022 68 0.067 31,108 14866 3516 90% 909 60,89 0.067 27,582 1,655 3516 75% 35157.35 0A67 25,876 1,553 3515 70% 795 53.B1 0.068 24,170 1,450 3516 60% 681 46,78 0.069 20,814 1,249 3516 50"% 568 39.57 0A70 17,573 1,054 3516 40% 454 32,47 0,072 14,445 367 3516 30°% 341 25.25 0.074 11,317 679 3516 25% 234 21.82 0.077 9,782 537 3516 20% 22713.47 0.031 9,246 495 3516 10% 11411.91 0.105 5,292 324 Electric Power Systems, Inc. 3305 Arctic Blvd Suite 201 ■ Anchorage, Alaska 99503 ■ Tel: (907) 522-1953 The last and third -to -last columns are calculated trivially from the other columns. Due to the limited access to performance data, these two performance data sets were mapped to all of the existing gensets as an approximation, even though each genset will have varying performance. This data will be used as described later in the analysis. The record data of power delivered to distribution is shown in excerpt below: Or Tags = 7 Sample Rate = 00:15:00 Old Feeders New Feeders Y Y Y [n [n N m U, x v 1 kQ z N k9 18 t8 a a a 20 20 20 �e ` 7/16/20 9:00 3,343.08 899.11 1,174,54 4,220.49 853.21 1,292.93 154.99 11,928.35 7/15/20 9:15 3,321.12 931.65 1,083.59 4,223.03 880.5 1,231,92 158.04 11,829.95 7/16/209:303,451.65 902.83 1,220.66 4,233,67 347.5 1,304.84 164.43 12,175.59 7/16/20 9:45 3,595.64 969.54 1,370.66 4,261.79 816.87 1,287.93 163,14 12,465.56 7/16/2010:00 3,560.39 870.07 1,406.89 4,282.75 741.14 1,292.71 155.18 12,319.13 7/16/2010:15 3,485,52 873.33 1,424.87 4,206.66 760.8 1,325.23 164.72 12,241.13 7/16/2010:30 3,521,94 392.59 1,472.79 4,207.49 769.85 1,357,46 165.35 12,387.46 7/16/2010:45 3,571.55 972.94 1,430.85 4,228,32 814.18 1,324.83 168.04 12,510.71 7/16/20 11:00 3,630,18 1,014,43 1,393.25 4,278.33 828.23 1,296.01 167.88 12,598.36 7/16/20IL15 3,618.05 1,012.71 1,429.34 4,243.01 830.2 1,317.83 173.45 12,529,60 7/16/2011:30 3,694.50 1,040.09 1,482.32 4,269.07 844.42 1,346.93 170.21 12,837.54 7/16/2011:45 3,637,54 1,049.47 1,428.28 4,249,90 851.3 1,319.60 170.94 12,707.03 7/16/2012:00 3,695.17 1,030.91 1,435.89 4,277.513 834.71 1,341.72 172.77 12,838.76 7/16/2012:15 3,646.80 962.55 1,421.69 4,301.20 787.15 1,295,10 172.39 12,585.89 7/16/2012:30 3,685.85 968.42 1,346.06 4,369.21 780.05 1,220,39 172.52 12,543.01 7/16/2012:45 3,765,70 939.71 1,411.67 4,415.76 747.83 1,247.00 174.5 12,702.17 7/16/2013:00 3,742.09 876.99 1,362.94 4,448.85 706.31 1,200.23 173.8 12,511.21 7/16/201115 3,784.31 921.65 1,451.20 4,418.75 733.96 1,269.24 181.75 12,750.76 7/16/2013:30 3,692,93 973.67 1,362.70 4,356.47 781,71 1,230.44 183.63 12,581.55 7/16/201145 3,644,51 926.76 1,301.45 4,363,61 748.34 1,187.29 187.22 12,359,19 7/15/2014:00 3,592.35 951.4 1,247.78 4,334.29 772.67 1,163,42 1,94.81 12,24&72 7/16/2014:15 3,587.55 925.29 1,225.33 4,352.06 750.51 1,140.43 196.4 12,167.57 7/16/2014:30 3,567.34 375.5 1,243.77 4,355.47 721,86 1,153,77 182.66 12,105,37 7/16/20 14:45 3,619.12 378.83.1,243.75 4,396.52 711,88 1,14109 185.43 12,172.62 7/16/2015:00 3,597,71 877.89 1,248.08 4,370.50 713.83 1,147.78 187.04 12,142.93 7/16/2015:15 3,602.97 872.99 1,281.42 4,360.30 711.29 1,194.60 195.23 12,208.80 7/16/2015:30 3,745.68 933.56 1,353.90 4,421.95 742.19 1,209.58 200.31 12,607.17 7/16/2015:45 3,641,83 906.11 1,320.43. 4,357.19 734.21 1,210.02 204.85 12,374.64 7/16/2016:00 3,689,84 925.76 1,296.03 4,402.88 739.E 1,177.25 200.04 12,431.40 7/16/2016:15 3,687.68 894.89 1,237.37 4,443.75 706.68 1,129.06 205.33 12,304.26 7/16/20 16:30 3,740.62 916.12 1,204.89 4,496.313 715.01 1,097.70 198.11 12,358.94 7/16/2016:45 3,594,46 383.34 1,132.20 4,418.61 707.94 1,068.41 202.47 12,007.43 7/16/2017:00 3,589.38 374.1 1,149.01 4,412.69 708.64 1,079.15 196.76 12,009.73 7/16/2017:15 3,588,79 869.63 1,133.22 4,393.34 704.75 1,105.58 202,33 12,048.69 7/16/20 17:30 3,689,99 925.4 1,269.51 4,416.61 733.13 1,157.51 195,41 12,389.56 7/16/2017:45 3,742.55 937.23 1,346.83 4,415.313 738.94 1,203.52 202.35 12,586.66 There are approximately 70,000 records, or 729 days of logged data. After filtering known bad data, approximately 53,750 records, or 560 days, remain. After this, the data is re -averaged from 15-minute windows to 6-hour windows. Due to the relatively consistent loading, this has minimal impact to the analysis results. Electric Power Systems, Inc. 3305 Arctic Blvd Suite 201 ■ Anchorage, Alaska 99503 ■ Tel: (907) 522-1953 ANALYSIS PROCEDURE For each genset, the prime kW nameplate capacity is procured: Unit 10 9 1 2 3 4 10 5 6 7 11 12A 12B 13 14 Simulate As 3516 35128 3512E 3512E 3512E 3516 3512E 3512E 3512E 3512E 3512E 3512E 3512B 35129 Prime 01 1,135 865 865 865 1,322 1,135 1,050 1,050 1,050 1.360 900 900 3,300 3,300 The nameplate capacity is then modified by a preferred capacity %, which is the % load that the operators prefer not to exceed for each given genset. This has minimal impact to the analysis results, but helps the data better match the real -world conditions. Unit Id 9 1 2 3 4 10 5 6 7 11 12A 12B 13 14 Simulate As 3516 3512E 3512B 35126 3512E 3516 3512B 3512E 3512E 3512E 3512E 3512E 3512E 3512E Prime kW 1,135 865 865 865 1,322 1,135 1,050 1,050 1,050 1.260 900 900 3.300 3,300 Preferred 76% 75% 75% 75% 70K 75% 72% 72% 72% 100% 10096 100% 100% 100% Pref. M 963 649 649 649 925 863 756 756 756 1,360 900 900 3,300 3,300 The gensets are then ordered by their stage order. This is the theoretical order that gensets are brought online as the load increases. The order is approximate, as it can be varied by the operators for a variety of reasons. The exact order will have minimal impact to the analysis results, but the staging order used was created based on our best understanding of the operator's preferrences, and is held constant across all records and scenarios to simplify the analysis. With the stage order developed, a capacity for each stage is developed. This is the sum of preferred kW for all gensets operating in that staging profile: Unit ID 9 1 2 3 4 10 5 6 7 :1 12A 12B 13 14 Simulate As 3515 3512E 3512E 3512B 3512E 3516 3512E 3512E 3512E 35128 3512E 3512E 3512E 3512E Prime M 1.135 865 865 865 1,322 1,135 1,050 1,050 1,050 1,360 900 900 3,300 3,3oo Preferred 7696 75% 75°% 75% 70M 76°% 72M 72°% 72M 100% 100% 100% 100% 100% Pref. kW 863 649 649 649 925 863 756 756 756 1,360 900 900 3,900 3,300 Staging Order 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Stage kw 863 1,511 2,160 2,809 3,734 4,597 5,353 6,109 6,865 8,225 9.125 10,025 13,325 16,625 This setup so far is identical for both the existing and future scenarios, but this next step is what separates the existing and future scenarios. The next step is to modify the stage capacity based on the scenario requirements to represent the capacity available to handle the distribution load. Where reserve spin is required, the stage kW previously calculated is modified by subtracting the largest single genset from the stage MW. In simple terms, the stage and scenario capacities are as follows: s stage kWS = Z genset Pref kWi i=1 scenario kW = stage kW - max(genset kW[1,...s]) Where s is the stage number, and subscript [1,... s] represents an array of gensets applicable to that stage. Electric Power Systems, Inc. 3305 Arctic Blvd Suite 201 ■ Anchorage, Alaska 99503 ■ Tel: (907) 522-1953 Unit I❑ 9 1 2 3 4 10 5 6 7 11 12A 12B 13 14 Simulate As 3516 35128 3512B 3512E 3512E 3516 3512B 3512E 3512E 351213 3512E 3512E 351213 3512E Prime M 1,135 865 8155 865 1,322 1,135 1,050 1,050 1,050 1,360 900 900 3,300 3,300 Preferred 76% 75% 75% 75% 7M 75% 72% 72% 72% 10096 100% 100% 100% 100% Pref. M 363 649 649 649 925 863 756 756 756 1,360 900 900 3,300 3,300 Staging Order 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Stage kw 963 1,511 2,160 2,809 3,734 4,597 5,353 6,1C9 6,865 8.225 9,125 10,025 13,325 15,625 Scenario M 0 649 1,298 1,946 2,809 3,671 4,427 5,1B3 5,939 6,865 7,765 8,665 10,025 13,325 For the future scenario, where no spin is required, the only change required is with the scenario kW. From before, the stage kW is calculated as: S stage kWS = Z genet pref kWi i=1 While the scenario kW is calculated by removed the largest genset to account for spin: scenario kWS = stage kWS - max(genset kW[1 s]) With the future scenario, the scenario kW does not have to be derated by the spin, and thus the full genset capacity is available, and thus the scenario kW is simply the stage kW: scenario kWS = stage kWS UnitIO 9 1 2 3 4 10 5 6 7 11 12A 12B 13 14 Simulate As 3516 3512E 3512E 3512E 3512E 3516 3512E 3512E 3512E 3512E 3512E 3512E 3512E 3512E Prime M 1,135 865 865 865 1,322 1,135 1,050 1,050 1,050 1,360 900 900 3,300 3,300 Preferred 7696 75% 759% 759% 70% 7696 72% 72% 729% 1009% 10096 100% 100% 1009% Pref. M 863 649 649 649 925 863 756 756 756 1,360 900 900 3,300 3,300 Staging Order 1 2 3 4 5 6 7 8 5 10 11 12 13 14 Stage kw 363 1,511 2,160 2,309 3,734 4,597 5,353 6,109 6,365 8,225 9,125 10,025 13,325 16,625 Scenario M 863 1,511 2,160 2,809 3,734 4,597 5,353 6,109 6,865 8,225 9,125 10,025 13,325 16,625 At this point, we have set up our genset and staging information developed, and now we can turn our attention to the record data. First, we calculate the number of generators needed to satisfy the load for each record. This is performed by taking the record load and "climbing" up the stage order to find a scenario kW capacity that will satisfy the record load. Next, the load % is determined by taking the record load and dividing it by the stage kW. Note that by determining the number of generators by the scenario kW, but determining the load by the stage kW, we are accounting for the lower generator loading due to the spin requirements. k g'- C Q E = G H I d K L M I N O P Q R 5 Unit 10 9 1 2 3 4 10 5 6 7 11 12A 12B 13 14 �■ Simulate As 3516 35120 3512B 3512B 3512B 3516 3512B 3512B 3512B 3512E 3512E 3512E 35128 35126 a Prime kW 1,135 865 855 865 1,322 1,135 1,050 1,050 1,050 1,360 900 900 3,300 3,300 4 1 Preferred 76% 75% 75% 75% 70% 75% 72% 72% 72% 100% 100% 100% 100% 100% 5 Pref. kW 863 649 649 649 925 863 756 756 756 1,360 900 900 3,300 3,300 5 Staging Order 1 2 3 4 5 6 7 8 9 10 11 12 13 14 7 Stage M 863 1,511 2,160 2,R09 3,734 4,597 5,353 6,109 6,865 8,225 13,125 10,025 13,325 15,625 . e Scenario kW 0 649 1,298 1,946 2,809 3,571. 4,427 5,183 5,939 5,865 7,765 9,665 10,025 13,325 12 (Record: ReckW Hours # Gens Load GAL GAL GAL GAL GAL GAL GAL GAL GAL GAL GAL GAL GAL GAL 13 11/15/1911 971.66 6.0 3 45% 169 142 142 ❑ 0 0 0 0 0 0 0 0 0 0 14 11/15/19 17 1603,49 6.0 4 57% 210 169 169 159 0 0 0 0 0 0 0 0 0 0 (spreadsheet excerpt showing the existing scenario and the first two records) Electric Power Systems, Inc. 3305 Arctic Blvd Suite 201 ■ Anchorage, Alaska 99503 ■ Tel: (907) 522-1953 For each staged generator (i) and each record (k), the fuel used is calculated by the following: genset GALik = GensetlnterpolateGPHperkWi(load%k * preferred%i) * load% * preferred kW * hours Where Gensetlnterpolate GPHperkWi is a custom function that interpolates the genset performance data based on the prime loading (load❑/ok * preferred%i) and provides the fuel performance in units of GPH of fuel per kW. This is then multiplied by the prime load (load❑/o * preferred kW.) to get GPH consumption of the genset at that record, and then multiplied by the record hours (hours k) to get the total gallons of fuel consumed during that record period. Between the existing and future scenarios, the calculations are the same, but the results change due to the difference in scenario kW between the two scenarios. See the following calculation excerpt for reference (records start on row 13; existing scenario are in columns E-J, and the future scenario are in columns W through AB): x 3 a 5 6 7 .a ix Retard: Rec M 13_ 11/15/1911 971.66 14 11/15/19 17 1.603,49 Unit ID Simulate As Prime M Preferred Pref. M Staging Order Stage M Scenario M Hours # Gens Load 6.0 3 4596 6.0 4 57% F G H I r 9 1 2 3 4 3516 35129 3512B 3512B 3512B 1,135 865 855 865 1,322 76% 75% 75% 75% 79% 863 649 649 549 925 1 2 3 4 5 863 1,511 2,150 2,809 3,734 0 649 1,298 1,945 2,909 GAL GAL GAL GAL GAL 169 142 142 ❑ ❑ 210 169 169 169 0 Unit ID 9 1 2 3 4 Simulate As 3516 3512E 3512E 3512E 3512E Prime M 1,135 865 865 865 1,322 Preferred 76% 75% 75% 75% 70% Pref. M 863 649 649 649 925 Staging Order 1 2 3 4 5 Stage kW 863 1,511 2,160 2,805 3,734 Scenario M 863 1,511 2,160 2,809 3,734 # Gens Load GAL GAL GAL GAL GAL 2 54% 233 184 0 ❑ ❑ 3 74% 265 206 206 ❑ ❑ (spreadsheet excerpt hiding all but the first 5 gensets and the first two records) Note how for the same record, the existing scenario has more operating gensets at a lower, less efficient, loading, whereas the future scenario has fewer gensets more efficiently loaded, due to the removal of spin requirements allowed by the BESS. The remaining calculations for run hours or heat rejection simply require replacing the genset GAL i,k calculation with the alternate calculations for run hours and heat rejection To develop per -month totals, the fuel consumption can be summed across the gensets by record, and then this subtotal can be summed across all records filtered by month. Due to the records representing more than one year, the resulting total represents more than one month of data. Thus the total is made uniform by multiplying by the ratio of the number of annual hours in the subject month to the sum of the record hours in the subject month. Eg, if the total gallons of fuel for all records in January was calculated as 51,858 gal, and the record data represented 1,488 hours of data for January, the 51,858 gal would be multiplied by the ratio of the number of annual hours in January (31 days * 24 hours = 744 hours) to the records hours: Jan fuel consumption = calculated record gallons * annual Jan hours / recorded Jan hours Jan fuel consumption = 51, 858 gal * 744 hours/1, 488 hours = 26, 793 gal The remaining calculations simply require replacing the genset GALik calculation with the alternate calculations for run hours and heat rejection. Electric Power Systems, Inc. 3305 Arctic Blvd Suite 201 ■ Anchorage, Alaska 99503 ■ Tel: (907) 522-1953 The tab "Data Summary" shows the summary of the longhand analysis noted above. Specifically, it extracts out the per -month results from the above, longhand calculation procedure. Additional calculations are there derived, including an approximate O&M cost savings estimate. ANALYSIS SUMMARY The following results are shown on the "Data Summary" tab. Fuel usage, runtime, JW heat rejection, and electrical production: FUELUSAGE RUNTIME Vr JW MMEH REJECTION ELECTRICALPRO DUCTION _ EXISTING FUTURE SAVINGS SAVINGS EXISTING FUTURE SAVINGS SAVINGS EXlMNG FUTURE REDUCTION REDUCTION (NOCHAII Month GAL GAL GAL % EXIST. HOURS HOURS HOURS % EXIST. MM6TH MMEITU MMRTU % EXIST. Peak kW Low kW kWh January 100,192 93,316 6,876 6.9% 3,201 2,241 960 30.0% 2,793 2,492 301 10.8% 2,243 1,404 1,350,262 February 86,022 91,058 4,924 5.736 2,735 1,995 740 27.0% 2,394 2,170 214 9.0% 2,103 1,458 1,169,917 March 96,915 91,486 5,429 5.6% 3,036 2,226 811 26.7% 2,691 2,444 236 9.8% 2,198 1,391 1,322,708 April 102,055 95.135 0,921 6.8% 3,180 2,253 527 29.2% 2,940 2,535 305 10.796 2,981 1,550 1,380,309 May 170,979 163,067 7,312 4.3% 4,464 3,542 822 18.4% 4,655 4,327 329 7.1% 7,404 695 2,359,410 June 176,363 170,502 5,867 3.3% 4,510 3,773 738 16.4% 4,795 4,531 264 5.5% 8,397 976 2,473,725 July 459,532 450.025 9,557 2.1% 8,336 8,058 828 9.3% 12,291 11,740 551 4.5% 12,334 3,584 6,5513,673 August 102,627 95,597 7,029 6.8% 3,246 2,244 1,002 30.9% 2,957 2,550 307 10.8% 4,019 1,321 1,386,500 September R0,563 73.722 6,840 8.5% 2,862 1,739 1,123 39.2% 2,268 1,976 291 12.9% 1,647 943 1,065,685 October 93,451 76,271) 7,181 8,5% 2,964 1,784 1,180 39.8% 2,349 2,043 306 13.0% 1,746 301 1,103,959 November 83,551 77,505 6,045 7.2% 2,857 1,B86 971 34.0% 2,339 2,079 259 11.1% 1,827 578 1,117,207 December 92,140 85,968 6,172 6.7% 3,036 2,098 948 31.2% 2,569 2,301 267 10.4% 2,130 1,348 1,240,841 Ye3r: 1,533,845 1,553,693 80,154 4.9% 44,977 33,929 11,048 248% 44,819 41,199 3,631 9.1% 49,130 15,640 22,540,197 O&M cost savings estimate (using the runtime hours savings): GENSET O&M SAVINGS D&M Record Costs: 2020 $422,216 2021 $641,203 2022 $573.063 Ave O&M/rar: $545,494 Runtime Savings: 24.6% O&M Savings: $133,992 Attachments: NEA BESS Scenario Analysis