RD 1 - 312 of 613

Attachment RML-RD-6 Page 5 of 13

2017 TX Rate Case

Southwestern Public Service Company Transmission Cost Recovery (TCRF)Baseline at June 30, 2017

40 Allocators Texas Jurisdiction 41 Transmission Demand 42 12CP-TRAN 46 58%

43 Production Demand 44 12CP-PROD 56 25%

42 Retail Transmission Demand 43 RETAIL-TRAN 71 81%

44 Transmission Plant in Service 45 PIS-TRAN 46 59%

46 Net Plant in Service 47 PIS-NET 52 99%

48 Direct Assigned 49 TX 100 00% 50 NM 0 00% 51 WHLS 0 00%

52 Transmission Transmission Radial 53 Interconnection System Lines 54 Transmission System (Functional) 55 DTRAN 0 000% 100 000% 0 000%

56 Transmission Radial Lines - Gross (Functional) 57 DTRANRADGRS 0 000% 0 000% 100.000%

58 Transmission Radial Lines - Depreciation (Functional) 59 DTRANRADDEP 0 000% 0.000% 100.000%

60 Transmission Radial Lines - Net (Functional) 61 DTRANRADNET 0 000% 0 000% 100 000%

62 Transmission Interconnection (Functional) 63 DPRODTI 100 000% 0 000% 0 000%

64 Plant in Service Transmission (Functional) 65 TRANPLT 1 419% 94 985% 3 152%

66 Plant in Service Net (Functional) 67 NETPLT 0 436% 40 844% 1 318%

RD 1 - 312 of 613 4300

Attachment RML-RD-6 Page 6 of 13

2017 TX Rate Case

Southwestern Public Service Company Transmission Cost Recovery (TCRF)Baseline at June 30, 2017

Customer Class Allocation ClassALLOC

68 Residential Service 33 552%

69 Small General Service 3 032%

70 Secondary General Service 18 856%

71 Primary General Service 11 245%

72 LGS-T 69 - 115 kV 5.576%

73 LGS-T 115 kV + 24 549%

74 Small Municipal and School Service 0.112%

75 Large Municipal Service 1 313%

76 Large School Service 1 500%

77 Municipal & State Street Lighting Service 0 155%

78 Guard & Flood Lighting Service 0 112%

79 Total 100 000%

RD 1 - 313 of 613 4301

£I9

JO t7

TE -

1 a

ll

Southwestern Public Service Company

Distribution Cost Recovery Factor (DCRF) Baseline

at June 30, 2017

Line No. Description

Distribution Costs

Substations Primary Ssstem

Secondary S rlem

Line Transormer

Sem ice Laterals LightinE Metering

Total Distribution

Costs

1 Distribution Insested Capital 2 Gross Plant m Sell ice 173,870,338 377,099,591 42,237,515 132,388,324 57,086,852 28,059,520 55,570,212 866,312,352

3 Accumulated Depmciation (41,033,702) (112,906,325) (12,588,634) (47,803,939) (23,302,926) (13,996,187) (23,774,983) (275,406,696)

4 Accumulated Deferred Income Taxes (35.205,799) (67,681,497) (7,604,959) (19,463,231) (6,370,400) (3,276,981) (6,139,094) (145,746,962)

5 Net Plant in Sen ice 97,630,837 196,511,769 22,043,922 65,116,154 27,413,526 10,786,351 25,656,134 445 158,694

6 Total Distribution InNestcd Capital - DIC,„ 97,630,837 196,511,769 22,043,922 65,116,154 27.413,526 10,786,351 25,656.134 445,158,694

7 Authorized Rate of Return on lmested Capital - RORco 7 91% 7 91% 7 91% 7 91% 7 91% 7 91% 7 91%

8 Return on Ins ested Capital 7,722,599 15,544,081 1,743,674 5,150,688 2,168.410 853,200 2,029,400 35,212,053

9 Approsed Distribution Charges

10 Depreciation Expense - DEPlintc 4,110,687 12,278,825 1,380,481 3,665,593 1,740,076 1,458,959 2,208,475 26,843,096

I I Property Tax 1,277,130 2,568,924 288,410 806,430 322,024 151,564 334,755 1,277,130

Other Taxes 5,484 11,031 1,238 3,463 1,383 651 1.437 5,484

12 Taxes Other Than Income Exel Pa) roll - COT„, 1,282,614 2,579,955 289,648 809,893 323,406 152,214 336,193 5,773,923

13 Income Tax Expense 14 Net Original Cost Rate Base 97,630,837 196,511,769 22,043,922 65,116,154 27.413,526 10,786,351 25,656,134 445,158,694

15 Return on Rate Base 7 91% 7 91% 7 91% 7 91% 7 91% 7 91% 7 91%

16 Eammgs 7,722,599 15,544,081 1,743,674 5,150,688 2,168,410 853,200 2,029,400 35,212,053

17 Net Onginal Cost Rate Base 97,630,837 196,511,769 22,043,922 65,116,154 27,413,526 10,786,351 25,656,134 445,158,694

18 Composite Cost of Debt 2 38% 2 38% 2 38% 2 38% 2 38% 2 38% 2 38%

19 Synehromzed Interest 2,323,614 4,676,980 524,645 1,549,764 652,442 256,715 610,616 10,594,777

20 Permanent Differences (455,976) 3,931 238 15,907 (418) 9,751 (19943) (446,508)

21 Taxable Income 4,943,010 10,871,032 1,219,267 3.616,831 1,515,550 606,236 1.398,841 24,170,768

22 Tax Rate 35 00% 35 00% 35 00% 35 00% 35 00% 35 00% 35 00%

23 Income Tax Expense 1,730,053 3,8(14,861 426,744 1,265,891 530,443 212,183 489,594 8,459,769

24 ITC Amortization 0

25 Subtotal Income Tax Expense 1,730,053 3,804,861 426,744 1,265,891 530,443 212,183 489,594 8,459,769

26 Income Tax Gross Up Factor 1 5527844 I 5527844 1 5527844 I 5527844 I 5527844 1 5527844 1 5527844

27 Total Income Tax Expense - FITor 2,686,400 5,908,129 662,641 1.965,655 823,663 329,474 760,234 13,136,197

Southwestern Distribution at June

28 29

30

Public Service Company Cost Recovery Factor (DCRF) Baseline

30, 2017

Calculation of ALLOC,,,,,, Term Customer Class

Distnbution Substations

NCP

Primary System NCP

Secondary System NCP

Line Transformers

NCP

Sen Replacement Lighting

Costs Direct-charged

Meter Replacement

Costs Total

Distribution ALLOCCLASS Residential Service 43 265% 43 265% 54 461% 54 461% 68 078% 59 102% 46 849%

31 NM Distribution Plant .12=,= $ 35 462 79U $ 18 662 575 42 239 771 $ 85 020 392 15,163,219 $ 208,554,046

32 Small General Son,. 4 359% 4 359% 5 487% 5 487% II 6570, II 993% 5363% 33 Nei Distribution Plant $ 1,209 506 $ 3,572,794 $ 3,195.689 $ 4 255 559 $ 8 565,607 $ 3.076,894 $ 23,876,047

34 Secondary General Service 25 551% 25 551% 32 163% 32 163% 17 640% 13 475% 25 043% 35 Net Distribution Plant $ 4,835.808 $ 24,945,695 , 50,210,802 , 014 .7.L.,90 $ 20 943.389 $ 3,457,171 $ 111,482,879

36 Primary General Service 20 036% 20 036% 10 569% 13 848% 37 Net Distribution Plant $ 19,560,894 , $ 39,372,251 $ 2,711,474 $ 61,644,618

38 Small Municipal and School Sen ice 0 207% 0 207% 0 261% 0 261% 1 113% 1 090% 0 319% 39 Net Distribution Plant 202 447 $ 407 486 $ 57.539 $ 169 966 $ 305 128 $ 279,576 $ 1,422,142

40 Large Municipal Sen ice 2 321% 2 321% 2 343% 2 343% 0 870% I 301% 2 121% 41 Net Distribution Plant $ 516,522 , $ 1,525,768 $ 2,266 432 $ 4,561,883 $ 238 483 $ 333,710 S 9,442.798

42 Largc School Service 3 394% 3 394% 4 194% 4 194% 0 641% I 452% 3 187% 43 Net Distnbution Plant $ 924,566 $ 2 731 102 $ 175 843 $ 3,313,946 , $ 6 670,324 $ 372,507 $ 14,188,287

0—. 44 Municipal & State Street Lighting 0 503% 0 503% 0 633% 0 633% 0 (100% 47 173% 0 000% 1 599%

I 45 Net Distribution Man( $ 491 154 $ 988 596 $ 139 595 $ 412 153 $ - $ 5 088 248 $ - $ 7,119,945 (..AJ

0 364% 0 364% 0 458% 0 458% 0 000% 52 827% 0 000% 1—k 46 Guard & Flood Lightmg 1 610% VI 47 Net Distribution Plant 941 .L...... ......

354 ....7.1.:.1,„428 881 .L.......,=.100 $ 297,994 $ - 1 ...

.:4211101 $ $ 7,166,347

0 1.•••+) 48 LGS-T 69 - 1151,V 0 184% 0 011%

Ch 1...., 14.3

49

50 51

Net Distribution Plant

LGS-T 115 *kV Net Distnbution Plant

$ 47,171 $ 47,171

0 048% 0 836% $ 214,411 $ 214,413

Southwestern

Distribution

at June

52 53 54 55

Public Service Company

Cost Recovery Factor (DCRF) Baseline

30, 2017

Calculation of DISTREV.„ ,. Term Customer Class Distnbution

Substations NCP

Pnmmy System NCP

Secondar) S!. stem NCP

Sem ice Replaceinent

Costs

Service Replacement

Costs

Meter Lighting Replacement

Direct-charged Costs DISTREVRC-CLASS 56 Residential Sen ice 43 265% 43 265% 54 461% 54 461% 68 078% 59 102% 57 DIC„ cLA„ $ 42,239,771 $ 85,020,392 $ 12,005,300 $ 35,462,790 $ 18,662,575 $ 15,163,219

58 ROR, 7 910% 7910% 7910% 7 910% 7 910% 7 910%

59 nir ___,x,s,.RORAT $ 3,341,166 $ 6,725,113 $ 949,619 $ 2,805,107 $ 1.476,210 $ 1,199,411

60 TIPPR --- Etc cuss $ 1,778,480 $ 5,312,407 $ 751,821 $ 1,996,312 $ 1,184,609 $ 1,305,247

61 F1T,, ,Ass $ 1,162,265 $ 2,556,139 $ 360,880 $ 1,070,512 $ 560,733 $ 449,312

62 OT, ,,,, $ 554,920 $ 1,116,212 $ 157,745 $ 441,074 $ 220.168 $ 198,696

63 D1C„ c,„ • RORAs + DEP13„ ,,,„ + FIT, c, ,. + OT,,,. $ 6,836,831 $ 15,709,872 $ 2.220,065 $ 6.313,005 $ 3,441,720 $ 3,152,665 $ 37,674,157

64 Small General Sen ice 4 359% 4 359% 5 487% 5 487% I I 65r/a 11 993%

65 DIC„ ri AsS $ 4,255,559 $ 8,565,607 $ 1,209,506 $ 3.572,794 $ 3,195,689 $ 3,07769,81904%

66 RORAT 7 910% 7 910% 7 910% 7 910% 7 910%

67 DIC „ ,,,,.. • ROR„ $ 336,615 $ 677.539 $ 95,672 $ 282,608 $ 252,779 $ 243,382

68 DEPR, MASS $ 179,178 $ 535,213 $ 75,744 $ 201,124 $ 202,847 $ 264,858

69 F1T,,,A. $ 117,096 $ 257,525 $ 36,358 $ 107,852 $ 96,017 $ 91,174

70 OT0( .,,,,. $ 55,907 $ 112,456 $ 15,892 $ 44,437 $ 37.701 S 40,319

71 DIC, crass • RORAT + DEPR,-, i /WS + Mar ,,,,,. + OT, CLAES $ 688,795 $ 1,582,733 $ 223,666 $ 636,021 $ 589,343 $ 639,733 $ 4,360,291

72 Secondary General Sol me

25 551% 25 551% 32 163% 32 163% 17 640% 13 475%

1-k 73 DIC„ ,Acs $ 24,945,695 $ 50,210,802 $ 7,090,014 $ 20,943,389 $ 4 835,808 $ 3'439171 74 ROR, 7 910% 7 910% 7 910% 7 910% 7 910% 7 % 0

I 75 DICitc.crass •ROR., $ 1,973,204 $ 3,971,674 $ 560,820 $ 1,656,622 $ 382,512 $ 273,462

UJ 76 DEPR0c „c„ $ 1.050,323 $ 3.137,368 $ 444,006 $ 1,178,969 S 306,953 $ 297,593 1-k

crN 77 78

FIT,,,c. OT ,A. „

$ 686,403

$ 327,721

$ 1,509,589

$ 659,205 $ 213,126

$ 93,160 $ 632,216 S 260,487

$ 145,296

$ 57,050 $ 04

$

1452;3402 2

0 •-1.) 79 DIC, c.,,, • ROR, + DEPR,,, A. + FIT, ,,,,,,s + OT, c, A. $ 4,037,652 $ 9,277,836 $ 1,311,112 $ 3,728,294 $ 891,812 S 718,799 $ 19,965,504

Ch 80 Pi imary General Service 20 036% 20 036% 10 569%

1-k 81 DIC, ,,,,A. $ 19,560,894 $ 39,372,251 $ 2,711,474

Co.) 82 ROR, 7 910% 7 910% 7 910%

83 DIC„ c, A., • RORAs $ 1,547,267 $ 3,114.345 $ 214,478

84 DEPR,,,,. $ 823,599 $ 2,460,133 $ 233,403

85 FIT,,,A. $ 538,236 $ 1,183,727 $ 80,346

86 OT„ c, A. $ 256,979 $ 516,909 $ 35,531

87 DIC„ a Ass • ROR, + DEPR„ „Ass + FIT„ csA. + OT, ,L,k. $ 3,166,081 $ 7,275,113 $ 563,757 $ 11,004,951

EI9

J0 L

IE -

1 W

I



M June 30, 2017

Distnbution Pnmary Secondary Sol ice &nice Meter 89 Substations Sy stem System Replacement Replacement Lighting Replacement 90 NCP NCP NCP Costs Costs Direct-charged Costs DISTREVRC-CLASS 91 Small Municipal and School Sell ice 0 207% 0 207% 0 261% 0 261% I 113% I 090% 92 $ 202,447 $ 407,486 $ 57.539 $ 169,966 $ 305,128 $ 279,576 93 RORAT 7 910% 7 910% 7 910% 7 910% 7 910% 7 910% 94 DIC RORAT $ 16,014 $ 32,232 $ 4,551 $ 13,444 $ 24,136 $ 22,114 95 DEPR,.eraas $ 8,524 $ 25,461 S 3,603 $ 9,568 $ 19,368 $ 24,066 96 FIT, cS.AS, $ 5,571 $ 12.251 $ 1,730 $ 5,131 $ 9,168 $ 8,284 97 OTrc ciass $ 2,660 $ 5,350 $ 756 $ 2.114 $ 3,600 $ 3.664 98 crAss • ROR,,, + DEPRa, rs Ass + FlT ANS + OTcc LLASS $ 32,768 $ 75,294 $ 10,640 $ 30,257 $ 56,271 $ 58,128 $ 263,359

99 Large Municipal Seta ice 2 321% 2 321% 2 343% 2 343% 0 870% I 301% 100 DIG. - - -,,, CI ASS $ 2,266,432 $ 4,561,883 $ 516,522 $ 1,525,768 $ 238,483 $ 333,710 101 ROA, 7910% 7 910% 7 910% 7910% 7 910% 7 910% 102 DICue.ci MS . RORAT $ 179,275 $ 360,845 $ 40,857 $ 120.688 $ 11,864 $ 26,396 103 DEPR, ,,,,,, $ 95,427 $ 285,044 $ 32.347 $ 85,890 S 15,138 S 28,726 104 FIT,,,A.s., $ 62,363 $ 137,153 $ 15.527 $ 46,058 $ 7,165 $ 9,888 105 OT, ,LAAS $ 29,775 $ 59.892 $ 6,787 $ 18,977 $ 2,813 $ 4.373 106 D1C,.ct Ass * RORAT + DEPR, c, Ass + Partc,Ass + OTreo ol ASS $ 366,839 $ 842,934 S 95,517 $ 271.614 $ 43,981 $ 69,383 $ 1,690,269

107 Large School Service 3 394% 3 394% 4 194% 4 194% 0 641% 1 452% 108 DIC„Ass $ 3,313,946 $ 6,670,324 $ 924,566 $ 2,731.102 $ 175,843 $ 372,507 109 RORA0 7 910% 7 910% 7 910% 7 910% 7 910% 7 910% 110 DICac-maaa • RORAr $ 262,133 $ 527,623 $ 73,133 S 216,030 $ 13,909 $ 29,465 I 1 1 DEPR„ ci aaa $ 139,532 S 416,788 $ 57,900 $ 153,742 $ 11,162 $ 32,065 112 FIT,,,,. $ 91,186 $ 200,543 $ 27,792 $ 82,444 $ 5,283 $ 11,038 113 OTRC CLAVI $ 43,537 $ 87,573 $ 12,148 $ 33,969 $ 2,074 $ 4,881 114 D1C, ccAss * ROR, + DEPRac-ccAss + MR/ CLASS + OT, CI ASS $ 536,388 $ 1,232,527 $ 170,974 $ 486,184 $ 32,429 $ 77,450 $ 2,535,952

115 Municipal & State Street Lighting 0 503% 0 503% 0 633% 0 633% 0 000% 47 173% 0 000% 116 D1C,,,s $ 491,154 $ 988,596 $ 139,595 $ 412,353 $ - $ 5,088,248 $ 117 RORAT 7910% 7910% 7 910% 7 910% 7910% 7 910% 7910% 118 D1C, (, A. • ROK, $ 38,850 $ 78,198 $ 11,042 $ 32.617 $ - $ 402,480 $ 119 DEPR,,...n,, S 20,680 $ 61,771 $ 8,742 $ 23,213 $ - $ 688.235 $ 120 FIT, CLAW $ 13,515 $ 29,722 $ 4,196 $ 12,448 $ $ 155,423 $ 121 OTS, Lusa $ 6,452 $ 12,979 $ 1,834 S 5,129 $ $ 71,804 $ 122 DICiu , i Ass • RORAT + DEPR, CLASS + FIT, rt Ass + DTI, crass $ 79,497 $ 182,671 $ 25,814 $ 73,406 $ $ 1,317,942 $ - $ 361,388

123 Guard & Flood Lighting 0 364% 0 364% 0 458% 0 458% 0 000% 52 82 rA, 0 000% 124 DIC,, ci ,,as $ 354,941 S 714.428 S l00,881 S 297,994 $ - S 5.698,103 S 125 ROR,,,, 7 910% 7 910% 7 910% 7 910% 7 910% 7 910% 7 910% 126 DIC,„ ,,,,,,• RORA , $ 28,076 $ 56,511 $ 7,980 $ 23,571 $ - $ 450.720 $ 127 DEPR0,.., us, $ 14,945 $ 44,640 $ 6,318 $ 16,775 $ - $ 770,724 $ 128 FIT,., ,,Ss $ 9,767 $ 21,479 $ 3.032 $ 8,996 $ $ 174,051 $ 129 OT, c,...,, $ 4,663 $ 9380 $ 1.326 $ 3,706 $ $ 80,410 $ 130 DIC„ , t „..• RORAT + DEPR„ ,,, Ass + FIT,,,,,, + OTR, r.,,,,,, $ 57.450 $ 132.010 $ 18,655 $ 53,048 $ $ 1,475,905 $ $ 261,164

Southwestern Public Service Company Distribution Cost Recovery Factor (DCRF) Baseline at June 30, 2017

131 132

LGS-T 69 - 115 kV DIC ASS, - C-LC

0 1S4% 47,171

133 ROR,, 7 910%

134 D1C, CLASS * RORAT 3,731

135 DEPR,,,,A. 4,060

136 FIT,,LAss 1,398

137 OTRCCL 618

138 DIC„,,Ass * RORAT + DEPRIec++Ass FITrtc, ASS + 0;1,1 MI 9,808 9,808

139 LGS-T 115 + kV 0 836%

140 DIC, CLASS 8 214,413

141 RORAT 7 910%

142 DIC„, I ASS • RORAT $ 16,960

143 DEPR„,,As, $ 18,457

144 FIT, CIA!. 8 6,353

145 OT,,,, A, $ 2,810

146 DIC,,c, A. • RORAT + DEPR„ ,,, A„ + F1T5 , „, + OT,,,,,,.. 8 44,580 $ 44,580

EI9

J0 8

TE- I

CM



at June 30, 2017

Texas 147 Allocators

Jurisdiction

148 Plant m Sun ice - Production, Transmission, Distnbution 149 PIS-PTD

150 Plant in Sen ice Retail - Production, Transmission. Distribution 151 PISRET-PTD

152 Plant m Sen - General 153 PIS-GEN

154 Demand - Non Coincident Peak - Distribution - TX Onb (kW) 155 NCP-DIST

156 Plant in Son ice - Distribution 157 PIS-DIST

158 Plant in Service - Net 159 PIS-NET

53 80%

70 72%

58 99%

99 94%

64 54%

52 99%

£I9

JO

6I E

- 1 C

DT 160 Direct Assigned

161 TX 162 NM 163 WH LS

100 00% 0 00% 0 00%

EI9

J0 O

a -

I

Southwestern Public Service Company Distribution Cost Recovery Factor (DCRF) Baseline at June 30, 2017

164 165 Allocators Substations

Pnmar) System

Secondary System

Line Transformers

Son ice Laterals Lighting Metering

166 Payroll Excludmg A&G (Functional) 167 SALWAGXAG 3 239% 10 416% 1 185% 1 128% 0 441% 2 593% 4 774%

168 Customer Accountmg (Emotional) 169 CUST 0 000% 0 000% 0 000% 0 000'4 0 000% 0 000% 0 000%

170 Distribution - Substations (Functional) 171 DDISPSUB 100 000% 0 000% 0 000%. 0 000% 0 000% 0 000% 0 000%

172 Distribution - Primary (Functional) 173 DD1STPOL / DDISTPUL 0 000% 100 000% 0 000% 0 000% 0 000% 0 000% 0 000%

174 Distribution - Secondary (Functional) 175 DDISTSOL / DDISTSUL 0 000% 0 000% 100 000% 0 000% 0 000% 0 000% 0 000%

176 Distribution - Line Transformers 177 DDISTSLT 0 000% 0 000% 0 000% 100 000% 0 000% 0 000% 0 000%

178 Distribution - Service Laterals (Functional) 179 CSERVICE 0 OM% 0 000% 0 000% 0 000% 100 000% 0 000% 0 000%

180 Distribution - Meters (Functional) 181 CMETERS 0 000% 0 000% 0 000% 0 000% 0 000% 0 WO% 100 000%

182 Distnbution - Installations on Customer Premises (Functional) 183 PLT_371 0 000% 0 000% 0 000% 0 000% 0 000% 100 000% 0 000%

184 Distribution - Street Lighting & Signal S) stems (Functional 185 PLT_373 0 000% 0 000% 0 000./0 0 000% 0 000% 100 000% 0 000%

186 Plant in Service - General (Functional) 187 GENLPLT 3 142% 10 106% I 149% I 095% 0 4270/0 2 516% 4 632%

188 Plant in Sen icc - Distribution (Functional; 189 D1STPLT 20 610% 43 801% 4 901% 16 047% 6 932% 2 778% 4 932%

190 Plant in Sen icc - Net (Functional) 191 NETPLT 5 619% I I 302% I 269% 3 548% 1 417%. 0 667% I 473%

Attachment RML-RD-7(CD) Page 1 of 1

2017 TX Rate Case


Workpapers of Richard M. Luth

2017 TX Rate Case

APPLICATION OF SOUTHWESTERN PUBLIC SERVICE COMPANY

FOR AUTHORITY TO CHANGE RATES

RML-RD-7(CD)

RD 1 - 321 of 613 4309

DOCKET NO.

APPLICATION OF SOUTHWESTERN § PUBLIC UTILITY COMMISSION PUBLIC SERVICE COMPANY FOR § AUTHORITY TO CHANGE RATES § OF TEXAS

DIRECT TESTIMONY of

JANNELL E. MARKS

on behalf of

SOUTHWESTERN PUBLIC SERVICE COMPANY

(Filename: MarksRDDirect.doc)

Table of Contents

GLOSSARY OF ACRONYMS AND DEFINED TERMS 2

LIST OF ATTACHMENTS 4

I. WITNESS IDENTIFICATION AND QUALIFICATIONS 5

II. ASSIGNMENT AND SUMMARY OF TESTIMONY AND RECOMMENDATION S 8

III. RATE FILING PACKAGE SCHEDULES 13 IV. LOAD RESEARCH 16

V. WEATHER'S EFFECT ON TEST YEAR SALES 22

VI. WEATHER'S EFFECT ON UPDATED TEST YEAR PEAK DEMAND 35

VII. FORECAST METHODOLGY 42

AFFIDAVIT 49

Marks Direct — Rate Design

Page 1

RD 1 - 322 of 613

4310

GLOSSARY OF ACRONYMS AND DEFINED TERMS

Acronym/Defined Term Meaning

Census Class Customer class in which all customers have IDR meters

Commission Public Utility Commission of Texas

DW Durbin-Watson

Golden Spread Golden Spread Electric Cooperative, Inc.

IDR Interval Demand Recorder

kW Kilowatt

kWh Kilowatt-hour

MW Megawatt

MWh Megawatt-hour

NCE New Century Energies, Inc.

NOAA National Oceanic and Atmospheric Administration

Non-census Class Customer class in which not all customers have IDRs

NSPM Northern States Power Company, a Minnesota corporation

NSPW Northern States Power Company, a Wisconsin corporation

Operating Companies NSPM, NSPW, PSCo, and SPS

PSCo Public Service Company of Colorado, a Colorado corporation

R2 statistic

Coefficient of determination

RFP

Rate Filing Package

Marks Direct — Rate Design Page 2

RD 1 - 323 of 613

4311

Acronvm/Defined Term Meaning

SPS Southwestern Public Service Company, a New Mexico corporation

Test Year April 1, 2016 through March 31, 2017

Update Period April 1, 2017 through June 30, 2017

Updated Test Year July 1, 2016 through June 30, 2017

Xcel Energy Xcel Energy Inc.


RD 1 - 324 of 613

4312

LIST OF ATTACHMENTS

Attachment Description

JEM-RD-1

Weather Normalization of Test Year and Updated Test Year Sales (Filename: JEM-RD-1.xlsx)

JEM-RD-2

Weather Normalization of Test Year and Updated Test Year Sales Wholesale and New Mexico (Filename: JEM-RD-2 .xl sx)

JEM-RD-3

Weather Normalization of Test Year and Updated Test Year Peak Demand (Filename: JEM-RD-3.xlsx)


Page 4

RD 1 - 325 of 613

4313

DIRECT TESTIMONY OF

JANNELL E. MARKS

1 I. WITNESS IDENTIFICATION AND QUALIFICATIONS

2 Q. Please state your name and business address.

3 A. My name is Janne11 E. Marks. My business address is 1800 Larimer Street,

4 Denver, Colorado 80202.

5 Q. On whose behalf are you testifying in this proceeding?

6 A. I am filing testimony on behalf of Southwestern Public Service Company, a New

7 Mexico corporation ("SPS") and wholly-owned electric utility subsidiary of Xcel

8 Energy Inc. ("Xcel Energy).

9 Q. By whom are you employed and in what position?

10 A. I am employed by Xcel Energy Services Inc., the service company subsidiary of

11 Xcel Energy, as Director of Sales, Energy and Demand Forecasting.

12 Q. Please briefly outline your responsibilities as Director of Sales, Energy and

13 Demand Forecasting.

14 A. I am responsible for the development of forecasted customer, sales, and peak

15 demand data and economic conditions for the Xcel Energy Operating Companies,

16 and for the presentation of this information to Xcel Energy's senior management,

17 other Xcel Energy departments, and various regulatory and reporting agencies. I

18 also am responsible for Xcel Energy's Load Research function, which designs,

19 maintains, monitors, and analyzes electric load research samples in the Xcel

20 Energy Operating Companies service territories. Finally, I am responsible for


Page 5

RD 1 - 326 of 613

4314

1 developing and implementing forecasting, planning, and load analysis studies for

2 regulatory proceedings.

3 Q. Please describe your educational background.

4 A. I graduated from Colorado State University with a Bachelor of Science degree in

5 Statistics.

6 Q. Please describe your professional experience.

7 A. I began my employment with Public Service Company of Colorado ("PSCo") in

8 1982 in the Economics and Forecasting Department. In 1985, I became a

9 Research Analyst, and, in 1991, I was promoted to Senior Research Analyst. In

10 that position, I was responsible for developing the customer and sales forecasts

11 for PSCo and the economic, customer, sales, and demand forecasts for Cheyenne

12 Light, Fuel and Power Company. In 1997, when PSCo merged with SPS to form

13 New Century Energies, Inc. ("NCE"), I assumed the position of Manager,

14 Demand, Energy and Customer Forecasts. In that position, I was responsible for

15 developing demand, energy, and customer forecasts for NCE's operating

16 companies, including SPS. I also directed the preparation of statistical reporting

17 for regulatory agencies and others regarding historical and forecasted reports. In

18 August 2000, following the merger of NCE and Northern States Power Company

19 that created Xcel Energy, I was named Manager, Energy Forecasting, with the

20 added responsibility for Northern States Power Company—Minnesota ("NSPM")

21 and Norther States Power Company—Wisconsin ("NSPW"). I assumed my

22 current position in February 2007, with the added responsibility for the Operating

23 Companies load research function.


Page 6

RD 1 - 327 of 613

4315

1 Q. Have you attended or taken any special courses or seminars relating to

2 public utilities?

3 A. Yes. I have attended the Institute for Professional Education's Economic

4 Modeling and Forecasting class and Itron's Load Forecasting Workshops. I have

5 also attended industry forecasting conferences and forecasting software user

6 group meetings and training classes sponsored by the Electric Power Research

7 Institute. I am a member of Itron's Energy Forecasting Group and Edison Electric

8 Institute's Load Forecasting Group.

9 Q. Have you testified before any regulatory authorities?

10 A. Yes. I have testified before the Public Utility Commission of Texas

11 (`CommissioC), the Colorado Public Utilities Commission, the Minnesota Public

12 Utilities Commission, the New Mexico Public Regulation Commission, the North

13 Dakota Public Service Commission, and the Public Service Commission of

14 Wisconsin on the issues of load research, sales and dernand forecasts, weather

15 normalization of sales and demand, and other related topics. I also have

16 submitted written testimony to the South Dakota Public Utilities Commission.


Page 7

RD 1 - 328 of 613

4316

1 II. ASSIGNMENT AND SUMMARY OF TESTIMONY AND

2 RECOMMENDATIONS

3 Q. What is your assignment in this proceeding?

4 A. The purpose of my testimony is to:

5 1. describe SPS's load research function and the load research information

6 that is used for cost allocation and rate design in this proceeding;

7 2. explain the methodology that SPS undertakes to measure normal weather

8 and to adjust both sales and demand that have been affected by abnormal

9 weather during the Updated Test Year (July 1, 2016 through June 30,

10 2017);1 and

11 3. discuss the process by which SPS forecasts information required for

12 Schedule 0-7.1 of the Rate Filing Package ("RFP").

13 In addition, I sponsor or co-sponsor the RFP schedules discussed in

14 Section III of this testimony, as well as the portions of the Executive Summary

15 that contain information from these schedules.

16 Q. Please provide a summary of conclusions and recommendations in your

1 7 testimony.

18 A. Load Research — Load research is the systematic collection and analysis of

19 customers electrical energy and demand requirements. SPS uses information

20 from Interval Demand Recorders (IDR")2 to determine the coincident and non-

21 coincident peaks for all customer classes. For the "Census classes," which are

22 customer classes in which all customers have IDRs, the IDR meters provide

23 actual measurements of demand. However, it is costly and not feasible to install

1 The Test Year in this case is the period from April 1, 2016 through March 31, 2017, and the Update Period is April 1, 2017 through June 30, 2017. The Updated Test Year consists of the last nine months of the Test Year and the three months in the Update Period.

2 IDRs are meters capable of recording loads for each interval of time.


Page 8

RD 1 - 329 of 613

4317

1 an IDR meter for every customer in every class. Therefore, for those customer

2 classes in which not all customers have IDRs, which are referred to as the "non-

3 Census classes," it is necessary to develop load research samples to estimate the

4 coincident and non-coincident peaks for the classes.

5 Using information from the IDR meters for the Census classes and

6 information from the load research samples for the non-Census classes, I have

7 provided various load research statistics to SPS witnesses Richard M. Luth and

8 Evan D. Evans, who incorporate those statistics in the class cost of service study

9 and rate design they present. Specifically, I provided the class coincident and

10 non-coincident peak demand for Census classes and the class coincident and non-

11 coincident load factors at peak for the non-Census classes. I recommend the

12 Commission approve those peak demands and load factors for purposes of

13 allocating costs among classes and designing rates.

14 Weather Normalization - SPS has calculated the effects of abnormal weather on

15 Updated Test Year sales. Consistent with the Commission's decisions in Docket

16 Nos. 404433 and 43695,4 SPS used a 10-year average to define normal weather.

17 Normal daily weather was based on the average of the last 10 years of historical

18 heating degree days, cooling degree days, and precipitation data. The Updated

19 Test Year heating degree days were 18.7% below normal; cooling degree days

20 were 7.0% above normal; and precipitation was 8.7% below normal. SPS

3 Application of Southwestern Electric Power Company for Authority to Change Rates and Reconcile Fuel Costs, Docket No. 40443, Order on Rehearing (Mar. 6, 2014).

4 Application of Southwestern Public Service Company for Authority to Change Rates, Docket

No. 43695 (Dec. 18, 2015).


RD 1 - 330 of 613 4318

1 calculated the effects of abnormal weather on Updated Test Year sales for

2 customer classes whose consumption patterns are affected by the weather using

3 weather normalization regression coefficients and econometric rnodels. The

4 overall adjustment was an increase of 2,817 megawatt-hours ("MWh") from the

5 Updated Test Year sales. This amounts to 0.02% of total Texas retail sales and is

6 the result of milder-than-normal winter weather being mostly offset by hotter- and

7 dryer-than-normal summer weather. Similarly, SPS also calculated the effects of

8 abnormal weather on the coincident peak demands in the Updated Test Year for

9 total retail and aggregated full requirements wholesale. Taken together, the

10 weather deviations resulted in an average of 4 megawatts ("MW") less retail peak

11 demand per month and an average of 3 MW less full requirement wholesale peak

12 demand per month from June through September of the Updated Test Year.

13 SPS also calculated the effect of abnormal weather on Golden Spread

14 Electric Cooperative, Inc.'s ("Golden Spread') full load peak demand coincident

15 with the SPS system peak demand. The average weather adjustment for the

16 Golden Spread full load peak demand coincident with the SPS system peak

17 demand for the four months of June, July, August, and Septernber of the Updated

18 Test Year was 4 MW per month.

19 I provided the MWh and MW impacts of abnormal weather to Mr. Luth,

20 who uses them to calculate present revenues and the allocation of production and

21 transmission capacity costs among classes.

22 I explain the methodology that SPS uses to weather normalize monthly

23 sales and demand amounts, as required by Schedule 0-2 of the RFP. SPS's


Page 10

RD 1 - 331 of 613

4319

1 weather-impacted sales are developed using industry standard regression

2 modeling techniques. SPS relies on a number of quantitative and qualitative tests

3 to ensure that its regression models are statistically valid. Thus, SPS's estimates

4 of weather normalized sales are reasonable and should be used to set rates in this

5 proceeding.

6 I recommend that the Commission approve the adjusted sales and demand

7 amounts resulting from the weather normalization discussed in this testimony.

8 Forecast Methodologv - I explain the methodology that SPS uses to

9 forecast the monthly sales and demand amounts, as required by Schedule 0-7.1 of

10 the RFP. Those monthly sales and demand forecasts are not used in the cost of

11 service, but instead are provided merely to comply with Schedule 0-7.1.

12 Q. Will your testimony and certain schedules you sponsor be updated to reflect

13 data for the period from April 1, 2017 through June 30, 2017, the Update

14 Period?

15 A. Yes. As explained by SPS witness William A. Grant, SPS will be using an

16 Updated Test Year in this case to determine its revenue requirement. Specifically,

17 in determining its proposed revenue requirement, SPS will replace the first three

18 months of the Test Year (April 2016 — June 2016) with the three months of the

19 "Update Period" (April 2017 — June 2017) to derive the "Updated Test Year."

20 The use of an Updated Test Year necessarily requires that certain costs provided

21 in SPS's Application will be based on estimates.

22 SPS will file an update no later than 45 days after filing its Application

23 that will replace the Update Period estimates with actual numbers. As part of


RD 1 - 332 of 613

4320

1 SPS's update filing, I will update my testimony to include the load research data

2 and calculations used for cost allocation and rate design. However, my testimony

3 provides the actual weather normalization adjustment for the Updated Test Year,

4 and, therefore, the weather-normalization testimony will not need to be updated. I

5 note that in the Update Filing Mr. Luth will update the calculations that affect

6 jurisdictional allocation, customer class cost allocation, and present revenues to

7 reflect the actual billing determinants for the Update Period. At that time, the

8 actual billing determinants for the Update Period will be adjusted by the weather

9 normalization amounts I have calculated for the Update Period.

10 Q. Were Attachments JEM-RD-1, JEM-RD-2 and JEM-RD-3 prepared by you

1 1 or under your direct supervision and control?

12 A. Yes.

13 Q. Were the portions of the RFP schedules and the portions of the Executive

14 Summary you sponsor or co-sponsor prepared by you or under your

15 supervision and control?

16 A. Yes.

17 Q. Do you incorporate the RFP Schedules and portions of the Executive

18 Summary sponsored or co-sponsored by you into this testimony?

19 A. Yes.


RD 1 - 333 of 613

4321

1 III. RATE FILING PACKAGE SCHEDULES

2 Q. Please list the RFP schedules that you sponsor or co-sponsor in this case.

3 A. I sponsor or co-sponsor the RFP schedules listed in Table JEM-RD-1:

4 Table JEM-RD-1

Schedule 0 1.3, 1.4, 1.9, 2.1(CD), 2.2(V)(CD), 2.3(V)(CD), 7.1, 8.1, 8.2, 8.3, 8.4, 9.1(CD), 9.2(V)(CD), 9.3(V)(CD), 10.1, and 10.2

Schedule Q 5.1, 5.2, and 5.3

5 Q. Please list the RFP schedules that you will update as part of SPS's update

6 filing?

7 A. As part of SPS's update filing, I will update the following schedules to reflect

8 data for the Updated Test Year (i.e., July 1, 2016 through June 30, 2017):

9 • Schedule 0 — 1.3, 1.4, and 1.9; and

10 • Schedule Q — 5.1 and 5.2.

11 Q. What information is contained in the 0-1 schedules that you sponsor?

12 A. The 0-1 schedules that I sponsor contain the following information:

13 • Schedule 0-1.3 contains unadjusted Test Year data by class for each

14 month of the Test Year and estimates for the Update Period for

15 coincident peaks at the source and at the meter, non-coincident peaks

16 at the source and at the meter, energy sales at the source, energy sales

17 by voltage level at the meter, and monthly class load factors and class

18 coincident peak load factors based on load research for the Test Year

19 and three previous years. I co-sponsor this schedule with Mr. Luth.

20 For this schedule, I sponsor the underlying load research used to

21 determine the coincident and non-coincident peaks.

22 • Schedule 0-1.4 contains adjusted Test Year data by class for each

23 month of the Test Year and estimates for the Update Period for

24 coincident peaks at the source and at the meter, non-coincident peaks

25 at the source and at the meter, energy sales at the source, energy sales

26 by voltage level at the meter, and monthly class load factors and class

27 coincident peak load factors based on load research for the Test Year

28 and three previous years. I co-sponsor this schedule with Mr. Luth.


Page 13

RD 1 - 334 of 613

4322

1 For this schedule, I sponsor the underlying load research and weather

2 normalization calculations used to determine the coincident and

3 non-coincident peaks presented.

4 • Schedule 0-1.9 contains total system and Texas retail peak demand by

5 class for the Test Year and for each month of the Test Year and

6 estimates for the Update Period. I co-sponsor this schedule with Mr.

7 Luth. For this schedule, I sponsor the underlying load research and

8 weather normalization calculations used in determining the peak

9 demands presented.

10 Q. What information is provided in the 0-2 schedules that you sponsor?

11 A. The 0-2 schedules contain information regarding the models used to derive

12 adjustments to the Test Year and Updated Test Year operating statistics provided

13 in Schedule 0-1. I explain the process by which I derived the adjustments in

14 Sections V and VI of this testimony.

15 Q. What information is provided in Schedule 0-7.1?

16 A. Schedule 0-7.1 contains the sales and demand forecasts for the rate year and 24

17 months following the rate year. I explain the process by which I derived the

18 forecasts in Section VII of this testimony.

19 Q. What information is contained in the 0-8 schedules?

20 A. The 0-8 schedules contain the information needed to perform a weather

21 normalization analysis:

22 • Schedule 0-8.1 contains 12 years of monthly weather data by weather

23 station, with calculations of actual heating degree days and cooling

24 degree days through the current time period.

25 • Schedule 0-8.2 contains the same information as is included in

26 Schedule 0-8.1, except the information has been weighted and

27 adjusted for billing cycles.

28 • Schedule 0-8.3 contains one year of normal heating degree days and

29 normal cooling degree days.


RD 1 - 335 of 613

4323

1

• Schedule 0-8.4 contains additional responses using a 65 degrees

2

Fahrenheit base temperature.

3 Q. What information is provided in the 0-9 schedules?

4 A. The 0-9 schedules contain information regarding the models used to derive the

5 sales and demand forecasts provided as part of Schedule 0-7.1.

6 Q. What information is provided in the 0-10 schedules you sponsor?

7 A. Schedule 0-10.1 contains 15 years of information on customer counts, revenues

8 from the sale of electricity, population, total employment, and total non-

9 agricultural employment. Schedule 0-10.2 provides 15 years of information on

10 nominal personal income and real personal income.

11 Q. You stated earlier that you also sponsor the Q-5 schedules. Please explain

12 what is included in those schedules.

13 A. The Q-5 schedules contain load research information for the Test Year and

14 estimated information for the Update Period. Schedule Q-5.1 contains the sum of

15 customer non-coincident maximum demand and class peak demand for the

16 Census classes, whereas Schedule Q-5.2 contains specified load research data for

17 non-Census classes. Schedule Q-5.3 contains a description of the method used to

18 develop demand estimates, including the sources of the data. I co-sponsor

19 Schedule Q-5.3 with Mr. Luth. For Schedule Q-5.3, I sponsor the description of

20 the load research methodology and data.


Page 15

RD 1 - 336 of 613

4324

1 IV. LOAD RESEARCH

2 Q. What is the purpose of load research?

3 A. Load research is the systematic collection and analysis of customers electrical

4 energy and demand requirements by time-of-day, month, season, and year. This

5 data, which includes load research samples, is collected and analyzed by customer

6 classes, stratums of customer classes, and other subsets of customer classes. Load

7 research enables utilities to better understand customers, their consurnption

8 patterns, their consumption responses to various factors, and the irnpact of

9 customers' energy requirements on the electric utility's system. In addition, load

10 research data is used to develop demand and energy allocators for cost allocation

11 studies and is used in designing rates.

12 Q. What are load research samples?

13 A. It is costly and not feasible to install IDR meters for all customers in all customer

14 classes. Therefore, it is necessary for SPS to develop load research samples to

15 determine the coincident and non-coincident peaks for certain classes. Load

16 research samples are subsets of the entire population that SPS surveys to estimate

17 the characteristics of the entire population. SPS's load research samples are

18 developed using a stratified random sampling method. This technique divides the

19 class of interest into smaller groups with like-characteristics. This method

20 effectively reduces the overall variance of the class, thereby reducing the sarnple

21 size. The samples are designed to meet or exceed the "90/10" load research


Page 16

RD 1 - 337 of 613

4325

1 standard specified by Federal Energy Regulatory Commission regulations

2 implementing the Public Utilities Regulatory Policies Act of 1978.5

3 Accuracy Level. If sample metering is required, the sampling rnethod and

4 procedures for collecting, processing, and analyzing the sample loads,

5 taken together, shall be designed so as to provide reasonably accurate data

6 consistent with available technology and equipment. An accuracy of plus

7 or minus 10 percent at the 90 percent confidence level shall be used as a

8 target for the measurement of group loads at the time of system and

9 customer group peaks. 10

11 Data validation is performed regularly on the load research samples to ensure that

12 the energy use of the sample corresponds closely with the population energy use.

13 Q. Does SPS use load research samples to determine the demand of all its

14 customer classes?

15 A. No. It is not necessary to conduct load research samples for customer classes in

16 which all customers have IDR meters because the IDR rneters provide actual

17 measurements of demand. Most of the customers with IDR meters are in the

18 Large General Service-Transmission class, although some Prirnary General

19 Service customers with on-site generation also have IDR meters. In addition, a

20 few of the customers with individual rate schedules have IDR meters installed.

21 As noted earlier, I refer to the classes in which all custorners have IDR meters as

22 "Census" classes. SPS uses the output of those IDR meters to determine the

23 Census classes demands for purposes of allocation, rate design, and billing.

5 Code of Federal Regulations, Title 18, Chapter 1, Subchapter K, Part 290.403, Subpart B.


RD 1 - 338 of 613

4326

1 Q. For which customer classes has SPS developed load research samples?

2 A. SPS develops load research samples for its non-Census classes throughout its

3 service territory in both Texas and New Mexico. SPS developed load research

4 samples for the following Texas retail customer classes:

5 • Residential Service;

6 • Residential Space Heating Service;6

7 • Large Municipal Service;

8 • Large School Service;

9 • Primary General Service;

10 • Secondary General Service;

11 • Small General Service; and

12 • Small Municipal and School Service.

13 Q. How does SPS go about performing the load research for the non-Census

14 classes?

15 A. Because it is cost-prohibitive to install an IDR meter for every customer, SPS

16 installs IDR meters on a random sample of customers in each non-Census class

17 (developed as I previously described). SPS then uses the electric usage data from

18 those sample customers to extrapolate the demand data for the remainder of the

19 class.

6 Residential Space Heating Service is an optional rider available through the Residential Service tariff. Although it is not a separate customer class, it is broken out for load research and weather adjustment purposes in my analysis and throughout my testimony. It is my understanding, however, that Mr. Luth combined Residential Service and Residential Space Heating Service into one class for the development of system coincident peak demands and class peak demands


Page 18

RD 1 - 339 of 613

4327

1 Q. What load research statistics did you provide for SPS's cost allocation study

2 and rate design?

3 A. For each SPS Census class, I provided the class coincident peak demand and non-

4 coincident peak demand. For each SPS non-Census customer class, I provided:

5 (1) the load factors at the time of the monthly system peak, which is the class

6 coincident peak; and (2) the load factors at the time of the monthly class peak,

7 which is the class non-coincident peak.

8 Q. Please define the terms "monthly system peak," "class coincident peak,"

9 "monthly class peak," and "class non-coincident peak."

10 A. The monthly system peak is the 60-minute interval in each month in which SPS's

11 system experiences the highest demand, and each class's demand during that

12 60-minute interval is the class coincident peak. The monthly class peak is the

13 30-minute interval in each month in which a class experiences its highest demand.

14 Unless the monthly class peak occurs during the same 60-minute interval as the

15 monthly system peak, the monthly class peak is a class non-coincident peak.

16 Q. What is a load factor?

17 A. A load factor is the ratio of the average load in kilowatts (kW") supplied during

18 a designated period to the peak or maximum load in kW occurring in that period.

19 For example, assume a customer used 10,000 kilowatt-hours (IWIC) during a

20 30-day period (720 hours) and had a maximum demand of 21 kW during this

21 same period. The customer's average load would be 13.89 kW (10,000 kWh /

22 720 hours = 13.89 kW). Dividing that number by 21 kW leads to 0.66 (13.89 / 21

23 = 0.66). That is then multiplied by 100% to arrive at a load factor of 66%.


RD 1 - 340 of 613

4328

1 Q. How did you determine each non-Census class's peak load factor?

2 A. I derived each non-Census class's system peak load factor from load research

3 samples.

4 Q. How did SPS use the non-Census class's load factors derived from your load

5 research and the Census class's peak demand data?

6 A. I provided the non-Census class coincident and non-coincident load factors at

7 peak and the Census class coincident and non-coincident peak demand for each

8 month to Mr. Luth who used them to develop demand allocators. Mr. Luth

9 discusses SPS's demand allocators in further detail in his testimony.

10 Q. How did SPS calculate the demand at the time of the monthly system peak

1 1 and the demand at the monthly class peak for the non-Census classes?

12 A. As explained by Mr. Luth, each non-Census class's demand at the time of the

13 system peak was calculated by applying the monthly system peak load factors

14 derived from the load research to the monthly sales by customer class. Each non-

15 Census class's demand at the time of the non-coincident peak was calculated by

16 applying the monthly class peak load factors derived from the load research to the

17 monthly energy sales by customer class.

18 Q. Did you make any adjustments to the class demands at the time of the

19 monthly system peaks?

20 A. Yes. Because the hourly loads for the sample classes are estimates, the sum of

21 each hourly demand, adjusted to generation level, will almost never equal SPS's

22 total system load. To account for this difference, the sample classes were

23 adjusted each month so that the sum of all hourly demand equals the hourly


Page 20

RD 1 - 341 of 613

4329

1 system load at the hour of SPS's monthly system peak demand. Mr. Luth

2 describes this process in his direct testimony. Both monthly system peak demand

3 by class and monthly non-coincident class peak demands were adjusted consistent

4 with the proportional allocation process discussed above.


Page 21

RD 1 - 342 of 613

4330

1 V. WEATHER'S EFFECT ON TEST YEAR SALES

2 Q. What topic do you discuss in this section of your testimony?

3 A. I explain the weather normalization that SPS perforrned to ensure that its Updated

4 Test Year sales and the present revenues calculated using those sales are adjusted

5 to eliminate the effects of abnormal weather.

6 Q. Did SPS calculate the effects on sales of abnormal weather for the Updated

7 Test Year?

8 A. Yes. Because the twelve months that comprise the Updated Test Year were hotter

9 and dryer than the 10-year average in SPS's service area during the cooling

10 season and warmer than the 10-year average during the heating season, SPS

11 calculated the effects of abnormal weather, as it has done in prior cases. The

12 Updated Test Year heating degree days were 18.7% below normal; the Updated

13 Test Year cooling degree days were 7.0% above normal; and the Updated Test

14 Year precipitation was 8.7% below normal. The percent difference from normal

15 is calculated using the following formula:

16 (Actual weather — Normal weather) / Normal weather

17 The calculation of the percent difference from normal weather is shown on page 1

18 of Attachment JEM-RD-1.

19 SPS calculated the effects on sales of abnormal weather during its

20 Updated Test Year for the following customer classes:


22 • Residential Space Heating Service;

23 • Small General Service;


Page 22

RD 1 - 343 of 613

4331


2

• Small Municipal and School Service;

3

• Large Municipal Service; and

4

• Large School Service.

5 SPS also weather normalized Updated Test Year sales for the Canadian River

6 Municipal Water Authority. SPS's research indicates that weather has little or no

7 effect on the consumption of the Primary General Service, Large General Service-

8 Transmission, and Street Lighting classes. Therefore, SPS did not make weather

9 adjustments for those classes.

10 Taken together the weather deviations resulted in 2,817 MWh less being

11 consumed in the Updated Test Year than would have been consumed in the

12 Updated Test Year with normal weather, which amounts to -0.02% of total Texas

13 retail sales. The calculation of the -0.02% appears on page 3 of Attachment

14 JEM-RD-1.

15 Q. How did SPS define the normal weather?

16 A. SPS used a 10-year average to define normal weather for purposes of this rate

17 case. Generally speaking, SPS agrees with the National Oceanic and

18 Atmospheric Administration's ("NOAA") view that normal weather should be

19 measured based on a 30-year period of time. But because the Commission

20 concluded in Docket No. 40443 that a 10-year period should be used to establish

21 normal values,' SPS calculated its weather adjustment in its most recent, fully-

7 Docket No. 40443, Order on Rehearing at 43-44, Finding of Fact Nos. 257-258 (Mar. 6, 2014).


RD 1 - 344 of 613

4332

1 litigated base rate case, Docket No. 43695, based on a 10-year normal period.8

2 SPS's weather normalization adjustment in Docket No. 43695 was approved by

3 the Commission, and SPS has taken the same approach in performing its weather

4 normalization adjustment in this case.

5 Q. What 10-year period did SPS use for weather normalization?

6 A. SPS used 120 months of actual weather data from January 1, 2006 through

7 December 31, 2015.

8 Q. Did SPS include the Updated Test Year in the 10-year period used to

9 calculate normal weather?

10 A. No. It is standard practice not to include the test year being normalized in the

11 calculation of normal weather. Using actual weather data from the 12-month

12 period used as the test year period in the calculation of the "normar weather may

13 create a bias toward the actual test year weather, which would potentially misstate

14 the variance of the test year weather from normal weather conditions. SPS has

15 applied this methodology for weather normalization adjustments in its past seven

16 rate cases, including Docket No. 43695. In Docket No. 43695, the Commission

17 adopted the Administrative Law Judge's determination that the factors included in

18 the calculation of normal weather should be independent of the test year weather

19 to which the normal weather is compared. In addition, NOAA also excludes the

8 SPS continues to use a 30-year definition of normal weather in preparing all of its internal

reporting, forecasting, and other management reporting. The other Xcel Energy Operating Companies—NSPM, NSPW, and PSCo—use different definitions of normal weather. Similarly to SPS, PSCo uses a 30-year period for defining normal weather in all of its internal reporting, forecasting, and management reporting. However, NSPM and NSPW have a long-standing practice of using a 20-year period for defining normal weather for internal reporting, forecasting, and management reporting.


RD 1 - 345 of 613

4333

1 current year's weather when calculating its 30-year normal weather statistics for

2 purposes of comparing and analyzing the weather for a particular month.

3 Q. How did SPS determine the normal weather?

4 A. Normal daily weather was based on the average of the last 10 years of historical

5 heating degree days, cooling degree days, and precipitation data used to develop

6 the weather adjustment coefficients for the Updated Test Year. The Updated Test

7 Year actual and normal cooling degree days, heating degree days, and

8 precipitation are reflected on page 1 of Attachment JEM-RD-1.

9 Q. What measure did SPS use to calculate heating degree days, cooling degree

10 days, and precipitation?

11 A. SPS used heating degree days and cooling degree days based on a 65-degree

12 Fahrenheit temperature base and rainfall equivalent precipitation measured in

13 inches as reported by NOAA for Amarillo and Lubbock, Texas. The weather data

14 is aggregated to the state level by weighting the individual weather station data by

15 the share of load in the Amarillo and Lubbock regions of the Texas service area.9

16 Q. Please explain how SPS calculated heating degree days.

17 A. SPS calculated heating degree days for each day by subtracting the average daily

18 temperature from 65 degrees Fahrenheit. For example, if the average daily

19 temperature was 45 degrees Fahrenheit, then 20 heating degree days were

20 calculated for that day. If the average daily temperature was greater than 65

21 degrees Fahrenheit, then that day recorded zero heating degree days. Daily

22 heating degree days are aggregated to monthly totals.

9 The weight for Amarillo is approximately 0.745, and the weight for Lubbock is approximately 0.255.


Page 25

RD 1 - 346 of 613

4334

1 Q. How did SPS calculate cooling degree days?

2 A. SPS calculated cooling degree days for each day by subtracting 65 degrees

3 Fahrenheit from the average daily temperature. For example, if the average daily

4 temperature was 75 degrees Fahrenheit, 10 cooling degree days were calculated

5 for that day. If the average daily temperature was less than 65 degrees Fahrenheit,

6 then that day recorded zero cooling degree days. Daily cooling degree days are

7 aggregated to monthly totals.

8 Q. Did the weather reflect the same billing days as the sales data?

9 A. Yes. To align the weather data with the same period of time as the billing-month

10 sales data, the heating degree days, cooling degree days, and precipitation data

11 were weighted by the number of times a particular day was included in a

12 particular billing month. These weighted heating degree days and cooling degree

13 days were divided by the total billing cycle days to arrive at average heating

14 degree days and cooling degree days for a billing month.

15 Q. How was the Updated Test Year weather adjustment calculated?

16 A. SPS calculated the weather adjustment using the deviation between normal and

17 actual weather, customer counts, and weather adjustment coefficients that

18 quantify the impact of a one-unit change in weather on sales per customer.

19 Q. How did SPS develop the weather adjustment coefficients used in the

20 weather normalization of sales?

21 A. SPS developed the billing-month coefficients for each weather-sensitive class

22 using econometric models.1° SPS then converted the billing-month coefficients to

10 An econometric model is a widely accepted modeling approach in which a linear regression

equation relates a dependent variable, such as sales, to a set of explanatory variables, such as economic and demographic concepts, customers, price, and weather. After the relationships are identified, forecasts of the explanatory variables can be used to predict future sales.


RD 1 - 347 of 613

4335

1 a calendar-month basis by prorating the modeled weather coefficients based on

2 the number of billing days in each billing month that occur in a particular calendar

3 month. Pages 12-17 of Attachment JEM-RD-1 reflect the conversion of modeled

4 weather coefficients to a calendar-month basis.

5 The data used in each of the models are:

6 • Historical billing-month sales by weather-sensitive class;

7 • Real personal income per household for the SPS Texas service

8 territory;

9 • Non-farm employment for the SPS Texas service territory;

10 • Weather (heating or cooling degree days);

11 • Seasonal binary variables;

12 • Precipitation variables;

13 • Customer counts;

14 • Number of billing days in each month;

15 • Population for the SPS Texas service territory;

16 • Other binary variables; and

17 • Autoregressive correction terms.

18 Q. How do the factors listed in the previous question affect sales?

19 A. Sales are expected to increase as each of the economic indicators increases and to

20 decrease as each economic indicator decreases. For example, if personal income

21 increases, electricity consumption will increase because customers have the

22 means to purchase and use more electricity-consuming products. Likewise, as

23 employment and population levels grow, electricity consumption is expected to

24 increase.


Page 27

RD 1 - 348 of 613

4336

1

Weather is also an independent variable that affects sales. The further the

2

average daily temperature deviates from 65 degrees Fahrenheit, the more cooling

3

degree days or heating degree days SPS will experience, which increases

4 electricity consumption. Similarly, SPS expects more sales to irrigation

5 customers when there is little precipitation, and it expects fewer sales to irrigation

6 customers when there is more precipitation.

7 Q. Please explain the difference between "billing-month" sales and "calendar-

8 month" sales.

9 A. SPS reads electric meters each working day according to a meter-reading

10 schedule based on 21 billing cycles per billing month. Meters read early in the

11 calendar month mostly reflect consumption that occurred during the previous

12 calendar month. Meters read late in the calendar month mostly reflect

13 consumption that occurred during the current calendar month. Consequently, the

14 "billing-montV sales for the current calendar month reflect consumption that

15 occurred in both the previous calendar month and the current calendar month.

16 Thus, billing-month sales lag calendar-month sales. In order to determine the

17 sales for a calendar month, SPS estimates "unbilled" sales, which is the electricity

18 consumed in the current calendar month that is not billed to the customer until the

19 succeeding calendar month.

20 Q. What is the purpose of estimating calendar-month sales?

21 A. Calendar-month sales are used to align the Test Year revenues with the relevant

22 Test Year expenses, which are reported on a calendar-month basis. SPS reflects

23 calendar-month revenue on its books for accounting and financial reporting

24 purposes.


Page 28

RD 1 - 349 of 613

4337

1 Q. Why is it necessary to convert the billing-month coefficients to calendar-

2 month coefficients?

3 A. Because the Updated Test Year sales being weather normalized are calendar-

4 month sales, the billing-month coefficients need to be converted to calendar-

5 month coefficients. After the billing-month coefficients are developed through

6 the econometric modeling process, the next step is to convert the billing-month

7 coefficients to a coefficient that represents a calendar month. SPS determines the

8 percentage of billing days for a calendar month that is billed in the current month

9 and that is billed in a future month. The monthly billing-month coefficient is

10 converted to a monthly calendar-month coefficient using these percentages.

11 Q. What was your source of economic and demographic data?

12 A. Historical economic and demographic variables for the counties in SPS service

13 territory, the state of Texas, and the nation were obtained from IHS Global

14 Insight, Inc., a source of data typically relied on by forecasting professionals. The

15 variables used in the models were service territory non-farm employment,

16 population, and real personal income per household. This information is used to

17 determine the historical relationship between sales and economic and

18 demographic measures.

19 Q. Please describe the regression models and associated analyses used in SPS's

20 weather normalization process.

21 A. The formulae in the regression models and associated statistics used in SPS's

22 weather normalization process are provided in Schedule 0-2.1 of the RFP.

23 Specifically, Schedule 0-2.1 shows, by customer class or major rate group, the


Page 29

RD 1 - 350 of 613

4338

1 formulae in the regression models with their summary statistics and descriptions

2 for each variable included in the model.

3 Q. What techniques did SPS employ to evaluate the validity of its regression

4 models?

5 A. There are a number of quantitative and qualitative validity tests that are applicable

6 to multiple regression analysis. Several of the more common tests SPS uses are

7 as follows:

8 First, the coefficient of determination (a2 statistic") test statistic is a

9 measure of the quality of the model's fit to the historical data. It represents the

10 proportion of the variation of the historical sales around their mean value that can

11 be attributed to the functional relationship between the historical sales and the

12 explanatory variables included in the model. If the R2 statistic is high, the set of

13 explanatory variables specified in the model are explaining a high degree of the

14 historical sales variability. All regression models used to develop the weather

15 normalization coefficients demonstrate R2 statistics larger than 86%, which is

16 satisfactory under this standard.

17 Second, the t-statistic of each variable indicates the degree of correlation

18 between that variable's data series and the sales data series being modeled. The

19 t-statistic is a measure of the statistical significance of each variable's individual

20 contribution to the prediction model. Generally, the absolute value of each

21 t-statistic should be greater than 1.960 to be considered statistically significant at

22 the 95% confidence level and greater than 1.645 to be considered statistically

23 significant at the 90% confidence level. This criterion was applied in the


RD 1 - 351 of 613

4339

1 development of the regression models used to develop the sales forecast. All

2 variables in the final regression models used to develop the weather normalization

3 coefficients tested satisfactorily under the 95% confidence level standard.

4 Third, each model was inspected for the presence of first-order

5 autocorrelation, as measured by the Durbin-Watson ("DW") test statistic.

6 Autocorrelation refers to the correlation of the model's error terms for different

7 time periods. For example, under the presence of first-order autocorrelation, an

8 overestimate in one time period is likely to lead to an overestimate in the

9 succeeding time period, and vice versa. Thus, when forecasting with a regression

10 model, absence of autocorrelation between the error terms is very important. The

11 DW test statistic ranges between 0 and 4, and provides a measure to test for

12 autocorrelation. In the absence of first-order autocorrelation, the DW test statistic

13 equals 2.0. Autoregressive correction terms were applied where appropriate so

14 that the final regression models used to develop the weather normalization

15 coefficients tested satisfactorily for the absence of first-order autocorrelation, as

16 measured by the DW test statistic.

17 Fourth, graphical inspection of each model's error terms (i.e., actual less

18 predicted) was used to verify that the models were not misspecified and that

19 statistical assumptions pertaining to constant variance among the residual terms

20 and their random distribution with respect to the predictor variables were not

21 violated. Analysis of each model's residuals indicated that the residuals were

22 homoscedastic (constant variance) and randomly distributed, indicating that the


Page 31

RD 1 - 3 52 of 613

4340

1 linear regression modeling technique was an appropriate selection for each

2 customer class sales that were statistically modeled.

3 Q. Please explain the steps you went through to complete the weather-

4 normalization calculation.

5 A. After calculating the calendar-month coefficients, I undertook a six-step process

6 to calculate the effect on sales of weather variance from normal conditions during

7 the Updated Test Year. The numbers used as examples in the six steps recounted

8 below appear on pages 4-5 of Attachment JEM-RD-1:

9 • Step 1 — I calculated the difference between the 10-year average

10 heating degree days in a particular month and the heating degree days

11 in that month of the Updated Test Year. For example, the 10-year

12 average number of heating degree days in October is 191, whereas the

13 number of heating degree days in October of the Updated Test Year

14 was 67, for a difference of approximately -124.

15 • Step 2 — I multiplied the difference calculated in Step 1 times the

16 number of customers in each class. For example, the Residential

17 Service class had 164,453 customers in October 2016, so I multiplied

18 -124 times 164,453, for a total of approximately -20,377,782.

19 • Step 3 — I then multiplied the result from Step 2 times the heating

20 degree day coefficient for that class to determine the number of MWh

21 resulting from the abnormal weather. Multiplying -20,377,782 times

22 the October 2016 coefficient for the Residential Service class, which is

23 0.0000016, yields -32 MWh.

24 • Step 4 — I then performed Steps 1-3 using the cooling degree data. For

25 October 2016, that calculation results in 5,288 MWh.

26 • Step 5 — I netted the heating degree MWh against the cooling degree

27 MWh for each class by month. That produces a total of 5,256 MWh

28 for the Residential Service class for October 2016 (-32 MWh + 5,288

29 MWh = 5,256 MWh).

30 • Step 6 — Finally, I totaled the number of MWh of all classes in each

31 month, and then I added the monthly amounts to arrive at the 12-

32 month total of -2,817 MWh attributable to abnormal weather. In other

33 words, actual weather resulted in total Test Year retail sales being

34 2,817 MWh lower than if weather had been normal.


RD 1 - 353 of 613

4341

1 Q. How did SPS use the weather-adjusted sales figures?

2 A. After calculating the weather-adjusted sales by class, I supplied those sales figures

3 to Mr. Luth, who used them to calculate present revenues. The numbers that I

4 provided to Mr. Luth are on page 4 of Attachment JEM-RD-1.

5 Q. Did SPS adjust its New Mexico retail sales during the Updated Test Year to

6 account for the effects of abnormal weather on New Mexico retail sales?

7 A. Yes. SPS adjusted the Updated Test Year sales for the weather-sensitive New

8 Mexico retail classes using the same process described for Texas retail sales.

9 These calculations are provided in Attachment JEM-RD-2. SPS relied on NOAA

10 weather data measured at weather stations in Roswell, New Mexico.

11 Q. Did SPS adjust its firm wholesale sales during the Updated Test Year to

12 account for the effects of abnormal weather on wholesale sales?

13 A. Yes. SPS adjusted the Updated Test Year sales for SPS firm wholesale customers

14 using weather adjustment coefficients developed for each wholesale customer and

15 weather specific to the location of each wholesale customer. The weather

16 adjustment coefficients for the wholesale customers were developed using

17 historical calendar-month sales for each customer, weather variables (heating or

18 cooling degree days or precipitation), and an economic indicator such as Gross

19 State Product. Since the coefficients are based on calendar-month sales, there is

20 no need to convert the coefficients from a billing-month basis to a calendar-month

21 basis. Sales to the wholesale customers in New Mexico were weather normalized

22 based on Roswell weather, and sales to wholesale customers in Texas were

23 weather normalized based on weather for either Amarillo or Lubbock. The


Page 33

RD 1 - 354 of 613

4342

1 calculations of the weather adjustment for firm wholesale sales are provided in

2 Attachment JEM-RD-2.

3 Q. Why does SPS adjust its New Mexico retail sales and its firm wholesale sales

4 for purposes of this Texas retail rate case?

5 A. Certain of the allocation factors SPS uses to allocate the components of the total

6 company cost of service among its three rate jurisdictions depend on relative

7 levels of sales for each jurisdiction. Consequently, to ensure that the allocation

8 percentages for each jurisdiction are determined on the same basis, it is necessary

9 to adjust the sales in all three jurisdictions to account for the effects of abnormal

10 weather on sales.

11 Q. Why does SPS use data from one weather station in New Mexico and two

12 weather stations in Texas?

13 A. SPS uses three weather stations because these three weather stations are

14 representative of SPS's service territory weather conditions. For example, based

15 on annual 2015 sales, 46.4% of SPS's weather-sensitive sales in Texas are to

16 customers located in Randall County and Potter County, which include and

17 surround Amarillo. Another 16.2% of SPS's weather-sensitive sales in Texas are

18 to customers located in the counties immediately surrounding Lubbock. In

19 addition, Roswell is a major population and economic center in the SPS New

20 Mexico service territory, and is close to the geographic center of the SPS New

21 Mexico service territory and, more specifically, close to the center of the weather

22 sensitive loads.


Page 34

RD 1 - 355 of 613

4343

1 VI. WEATHEWS EFFECT ON UPDATED TEST YEAR PEAK DEMAND

2 Q. What topic do you discuss in this section of your testimony?

3 A. I explain how SPS calculated the effects of abnormal weather on coincident peak

4 demands in the Updated Test Year.11 For the same reasons I explained in Section

5 V of my testimony.

6 Q. Did SPS calculate the effects of abnormal weather on its Updated Test Year

7 system peak demand?

8 A. Yes. Because weather varied from normal during the Updated Test Year, it was

9 necessary to adjust the Updated Test Year coincident peak demand to account for

10 weather for the following customer groups:

11

• Total retail; and

12

• Aggregated full requirements wholesale.

13 For the same reason I explained in Section V of my testimony, I adjusted the peak

14 demands in all three of SPS's rate jurisdictions to ensure that the allocation

15 percentages for each jurisdiction are determined on the same basis for purposes of

16 this rate case.

17 Q. What source of weather did SPS use to measure the adjustment?

18 A. SPS used a combination of peak day average daily temperature, peak day heating

19 degree days, and accumulated precipitation for the week prior to the peak day to

20 measure weather adjustments for peak demand. These weather values were

21 calculated using weather data reported from the NOAA weather stations in

22 Amarillo, Lubbock, and Roswell. The total SPS weather is an average of the

1 1 SPS does not weather-normalize non-coincident peak demands.


Page 35

RD 1 - 356 of 613

4344

1 Amarillo, Lubbock, and Roswell weather station data weighted by sales

2 associated with the respective regions of the SPS service area.

3 Q. How did SPS calculate average peak day temperature?

4 A. The peak day average temperature was calculated by adding the peak day

5 maximum daily temperature and peak day minimum daily temperature, and then

6 by dividing that amount by 2. For example, if the peak day maximum

7 temperature was 55 degrees Fahrenheit and the peak day minimum temperature

8 was 35 degrees Fahrenheit, the average peak day temperature would be 45

9 degrees Fahrenheit.

10 Q. Please explain how SPS calculated the peak day heating degree days.

1 1 A. SPS calculated peak day heating degree days by subtracting the peak day average

12 temperature from 65 degrees Fahrenheit. For example, if the peak day average

13 daily temperature was 45 degrees Fahrenheit, then 20 heating degree days were

14 calculated for that day. If the average peak day temperature was greater than 65

15 degrees Fahrenheit, then that peak day recorded zero heating degree days.

16 Q. How did SPS calculate precipitation?

17 A. SPS calculated the accumulation of water-equivalent precipitation for the seven

18 days prior to the peak day, measured in inches.

19 Q. How did SPS define the normal weather?

20 A. As noted earlier, SPS agrees with NOAA's definition that normal weather is

21 representative of typical weather based on a 30-year period. However, given the

22 Commission's ruling in favor of using a 10-year period to measure normal

23 weather in Docket No. 40443 and SPS's most recent, fully-litigated base rate


Page 36

RD 1 - 357 of 613

4345

1 case, Docket No. 43695, SPS calculated its weather adjustment based on a

2 10-year normal period for this filing.

3 Q. How did SPS determine the normal weather?

4 A. Normal peak day weather was based on the average of the 10-year period from

5 January 2006 to December 2015 for the peak day of each month for historical

6 average daily temperature, heating degree days, and precipitation. The Updated

7 Test Year and normal-weather data for maximum temperatures, heating degree

8 days, and precipitation are summarized on page 1 of Attachment JEM-RD-3.

9 Q. How was the Updated Test Year weather adjustment calculated?

10 A. SPS calculated the peak demand weather adjustment using the deviation between

11 normal and actual weather and weather adjustment coefficients that quantify the

12 impact of a one-unit change in weather on retail and full requirements wholesale

13 peak demand.

14 Q. How did SPS calculate the coefficients used in the peak demand weather

15 normalization calculations?

16 A. SPS developed the peak demand weather coefficients for the retail coincident

17 peak demand and full requirements wholesale coincident peak demand using

18 econometric models. The data used in the models include historical peak demand

19 and sales for each customer group, as well as the weather concept variables

20 (average temperature, heating degree days, and precipitation for the week prior to

21 the peak day). Each regression model also has an autoregressive error correction

22 term variable. The regression models and associated statistics are provided in

23 Schedule 0-2.1 of the RFP.


Page 37

RD 1 - 358 of 613

4346

1 Q. What dependent variables does SPS use in the regression models?

2 A. The dependent variables used to develop peak demand weather coefficients are

3 the monthly coincident peak demands for each customer class. The first

4

explanatory variable in each regression rnodel is a 12-month moving average for

5

the respective sales for each customer class. The next set of explanatory variables

6

in each regression model uses the following weather concept variables:

7

• Average peak day temperature;

8

• Peak day heating degree days; and

9

• The accumulation of precipitation for the seven days prior to the peak

10

day of each month.

11 Q. How did the Updated Test Year peak day weather for the June through

12 September period compare to normal weather?

13 A. The total SPS Updated Test Year surnmer months (July through Septernber, 2016

14 and June 2017) peak day average daily ternperature was 0.7% above normal, and

15 accumulated precipitation was 89.2% above normal. As shown on Page 2 of

16 Attachment JEM-RD-3, taken together these weather deviations resulted in an

17 average of 4 MW, or 0.1%, less retail peak demand per month and an average of

18 3 MW, or 0.3%, less full requirement wholesale peak demand per month from

19 June, July, August, and September in the Updated Test Year compared to normal

20 weather. SPS adjusted the Updated Test Year peak demand for deviations of the

21 actual Updated Test Year weather from the 10-year average weather.


Page 38

RD 1 - 359 of 613

4347

1 Q. Please explain the steps you went through to complete the peak demand

2 weather-normalization calculation.

3 A. I undertook a four-step process to calculate the effect on peak demand of weather

4 variance from the 10-year normal conditions during the Updated Test Year. The

5 numbers used as examples in the four steps recounted below are for the retail peak

6 demand and appear on page 3 of Attachment JEM-RD-3:

7 • Step 1 — I calculated the difference between: (i) the 10-year average

8 weather concepts (as measured in average peak day temperature, peak day

9 heating degree days, and precipitation) in a particular month, and (ii) the

10 actual weather concept in that month of the Updated Test Year. For

11 example, for the retail peak demand the 10-year average peak day

12 temperature in July is 84.1 degrees Fahrenheit, whereas the actual average

13 peak day temperature in July of the Test Year was 90.3 degrees

14 Fahrenheit, a difference of 6.2 degrees Fahrenheit. This step is repeated

15 for each weather concept. The 10-year average of precipitation for the

16 week preceding the peak day in July is 0.32 of an inch of precipitation,

17 whereas the actual precipitation for the week preceding the peak day in

18 July of the Updated Test Year was 0.06 of an inch, for a difference of -

19 0.26 of an inch of precipitation. The 10-year average of heating degree

20 days on the peak day in July is 0.0, and the actual peak day heating degree

21 days in July of the Updated Test Year was also 0.0, resulting in no

22 difference from normal.

23 • Step 2 — The variance in weather from the 10-year average from Step 1 for

24 each weather concept is multiplied by the respective weather adjustment

25 coefficient to determine the number of MW resulting from the variance in

26 actual weather from the 10-year average weather. Weather adjustment

27 coefficients are developed with econometric models using the same

28 methodology described earlier. To continue with the retail peak demand

29 example from Step 1, multiplying the variance in average peak day

30 temperature of 6.2 degrees Fahrenheit times the July 2016 coefficient for

31 average peak day temperature, which is 11.1902, yields 69 MW. This step

32 is repeated for each weather concept. Multiplying the variance in

33 precipitation one week prior to the peak day of -0.26 inches times the July

34 2016 coefficient for precipitation one week prior to the peak day, which is

35 -65.30827, yields 17 MW. Because there is no weather adjustment

36 coefficient in July for heating degree days, this weather concept did not

37 have a weather adjustment in July.


RD 1 - 360 of 613

4348

1 • Step 3 — For each month, I summed the weather adjustments calculated in

2 Step 2 from each weather concept. Continuing with the example from

3 Step 1 and Step 2, this step produces a total weather adjustment of 86 MW

4 for the total retail peak demand for July 2016 69 MW + 17 MW =

5 86 MW).

6 • Step 4 — Finally, I averaged the weather adjusted MW for the summer

7 months of the Updated Test Year for June 2017, July 2016, August 2016,

8 and September 2016 to arrive at a 4-month average of weather's impact on

9 the peak demand. Continuing with the example from the previous steps,

10 the average weather adjustment for the retail peak demand for the four

11 months of June, July, August, and September of the Updated Test Year

I 2 was -4 MW per month. Using the same methodology described in Step 1

13 through Step 4, the average weather adjustment for the full requirement

14 wholesale peak demand for the four months of June, July, August, and

15 September of the Updated Test Year was -3 MW per month. Pages 3 and

16 4 of Attachment JEM-RD-3 contain the weather adjustment calculations

17 for the retail peak demand and the full requirements wholesale peak

18 demand, respectively.

19 Q. Did SPS calculate the effect of weather on the Golden Spread full load peak

20 demand coincident with the SPS system peak demand?

21 A. Yes. I calculated the effect of weather on the Golden Spread full load peak

22 demand coincident with the SPS system peak demand using the same

23 methodology previously described for weather adjusting the retail and full

24 requirement wholesale peak demand. The weather values used to adjust the

25 Golden Spread full load peak demand were calculated for the Texas Panhandle

26 region. As I explained earlier, the Texas Panhandle weather is an average of the

27 Amarillo and Lubbock weather station data weighted by sales associated with the

28 respective regions of the SPS service area located in Texas. The peak demand

29 weather coefficients for Golden Spread were developed using an econometric

30 rnodel. The data used in the model includes historical peak demand for Golden

31 Spread, as well as the weather concept variables for average temperature and


Page 40

RD 1 - 361 of 613

4349

1 accumulated precipitation for the week prior to the peak day, and an

2 autoregressive error correction term variable. For the Texas Panhandle, the

3 Updated Test Year summer months (June, 2017 and July, August, and September,

4 2016) peak day average daily temperature was 2.0% above normal, and the

5 accumulated precipitation was 99.4% above normal.

6 Q. What is the weather adjustment for the Golden Spread full load peak

7 demand coincident with the SPS system peak demand for the Updated Test

8 Year?

9 A. As shown on Page 2 of Attachment JEM-RD-3, the average weather adjustment

10 for the Golden Spread full load peak demand coincident with the SPS system

11 peak demand for the four months of June, 2017 and July, August, and September,

12 2016 of the Updated Test Year was 4 MW per month. Page 5 of Attachment

13 JEM-RD-3 provides the weather adjustment calculation for the Golden Spread

14 full load peak demand.

15 Q. How did SPS use the weather-adjusted peak demand figures?

16 A. After calculating the weather adjustments for peak demand by customer class, I

17 supplied those peak demand adjustments to Mr. Luth, who used them to calculate

18 the allocation of production and transmission capacity costs among classes. The

19 numbers that I provided to Mr. Luth are on page 2 of Attachment JEM-RD-3.


Page 41

RD 1 - 362 of 613

4350

1 VII. FORECAST METHODOLGY

2 Q. What topic do you discuss in this section of your pre-filed testimony?

3 A. I describe the process that SPS undertakes to derive the rate year sales and

4 demand forecasts as required by Schedule 0-7.1 of the RFP. However, SPS does

5 not use the forecasted information discussed in this section of my testimony to set

6 rates. The forecasting is done solely for the purpose of complying with the

7 Schedule 0-7.1 requirements.

8 Q. Please briefly describe SPS's methodology for forecasting rate year energy

9 and demand values.

10 A. SPS forecasts monthly customer counts, retail sales, retail peak demand, and full

11 requirements wholesale sales using econometric forecasting models.

12 Q. What inputs does SPS use in its econometric forecasting models?

13 A. The inputs used to arrive at the dependent variable in the econometric forecasting

14 models are: (1) SPS's historical customer counts and billing-month retail MWh

15 sales by jurisdiction and class; (2) SPS's historical calendar-month MWh sales for

16 each full requirements wholesale customer; and (3) SPS's historical retail peak

17 demand. For the explanatory variables, SPS obtains historical and forecasted

18 economic and demographic variables for the nation, state, and SPS service

19 territory from IHS Global Insight, Inc. Those explanatory variables are:

20 • population;

21 • nurnber of households;

22 • number of customers;

23 • Gross County Product;

24 • Gross State Product;


Page 42

RD 1 - 363 of 613

4351

1

• Gross Domestic Product;

2

• Consumer Price Index;

3

• employment;

4

• personal income;

5

• electricity prices;

6

• oil prices;

7

• Industrial Production index for oil and gas extraction.

8

SPS also obtains weather information to use as explanatory variables in the

9 models. As discussed earlier, that weather information, which comes from

10 NOAA weather stations at the Amarillo International Airport located in Amarillo,

11 Texas, at the Lubbock Regional Airport located in Lubbock, Texas, and at the

12 Roswell Industrial Air Center located in Roswell, New Mexico, is composed of

13 65 degree based heating degree days, 65 degree based cooling degree days,

14 precipitation, and temperature. Other variables used by SPS include

15 autoregressive and moving average correction terms, the number of days in the

16 calendar month and the billing month, total retail sales, seasonal binary variables,

17 trend variables, and various other binary variables.

18 Q. What are the outputs from SPS's econometric models?

19 A. The outputs from the econometric models are forecasts of customer counts,

20 billing-month retail MWh sales by jurisdiction and class, calendar-month MWh

21 sales for each full requirements wholesale customer, and retail peak demand.


Page 43

RD 1 - 364 of 613

4352

1 Q. What techniques does SPS use to evaluate the validity of its quantitative

2 forecasting models and sales projections?

3 A. Similar to the weather normalization models, SPS uses the R2 statistic, t-statistics,

4 DW test statistic, and graphical inspection of each model's error terms to evaluate

5 the validity of its models and projections.

6 The linear regression models used to develop SPS's sales forecasts

7 produce high R2 statistics, ranging between 0.691 and 0.991. The linear

8 regression models used to develop SPS's customer forecasts produce similarly

9 high R2 statistics, ranging between 0.998 and 0.999. The t-statistics of the

10 explanatory variables were statistically significant at the 95% confidence level

11 except for two variables which were statistically significant at the 90% confidence

12 level. Autoregressive correction terms were applied where appropriate so that the

13 final models used to develop the sales forecast tested satisfactorily for the absence

14 of first-order autocorrelation, as measured by the DW test statistic.

15 Additionally, analysis of each model's residuals indicated that the

16 residuals were homoscedastic (of constant variance) and randomly distributed,

17 which demonstrated that the linear regression modeling technique was an

18 appropriate selection for each customer class's sales that were statistically

19 modeled.

20 Finally, SPS reviewed the statistically modeled sales forecasts for each

21 customer class for reasonableness as compared to the respective monthly sales

22 history for that class. Graphical inspection reveals that the forecast patterns fit

23 well with the respective historical patterns for each customer class. The annual


RD 1 - 3 6 5 of 613

4353

1 total forecasted sales have been compared to their respective historical trends for

2 consistency. The forecast models and model data are provided in Schedules

3 0-9.1(CD) through 0-9.3(V)(CD).

4 Q. Did SPS make any adjustments to the outputs?

5 A. Yes. SPS converted the billing-month sales data to calendar-month sales data. In

6 addition, SPS adjusted the model output for incremental Demand Side

7 Management savings, as well as load growth or load reductions that are identified

8 by commercial customer account managers and that would not be captured in the

9 historical modeling data.

10 Q. How did SPS determine the estimated monthly calendar-month sales for the

11 forecast period?

12 A. SPS calculated the calendar-month sales based on the projected billing-month

13 sales for the following customer classes:


15 • Residential Space Heating Service;

16 • Small General Service;


18 • Irrigation Service;

19 • Primary General Service;

20 • Large General Service - Transmission;

21 • Large Municipal Service;

22 • Large School Service;

23 • Small Municipal and School Service; and

24 • Street and Area Lighting.


Page 45

RD 1 - 366 of 613

4354

1 SPS calculated the calendar-month sales both in terms of the sales load

2 component that is not associated with weather (i.e., "base load") and the sales

3 load component that is influenced by weather (i.e., "total weather load"). The

4 weather was measured in terms of normal heating degree days, cooling degree

5 days, and precipitation as described earlier. SPS calculated the base load sales

6 and the total weather load sales components for each class, and the two

7 components were then combined to provide the total calendar-month volumes.

8 Q. How did SPS calculate the calendar-month base load component?

9 A. SPS calculated the calendar-month base load cornponent using four steps:

10 • Step 1 — SPS calculated the billing-month total weather load by

11 multiplying the billing-month sales weather normalization regression

12 coefficients (defined in terms of billing-month heating degree days,

13 cooling degree days, and number of customers), times billing-month

14 normal heating degree days and cooling degree days, and then multiplying

15 the product times the projected customers.

16 • Step 2 — SPS calculated the billing-month base load sales by taking the

17 difference between the projected total billing-month sales and the

18 billing-rnonth total weather load (as calculated in Step 1).

19 • Step 3 — SPS next deterrnined the billing-month base load sales per billing

20 day by dividing the billing-month base load sales (from Step 2) by the

21 average nurnber of billing days per billing month.

22 • Step 4 — SPS then calculated the calendar-month base load sales by

23 multiplying the billing-month base load sales per billing day (from Step 3)

24 times the number of days in the calendar month.

25 Q. How did SPS calculate the calendar-month total weather load component?

26 A. SPS calculated the calendar-rnonth total weather load component in the sarne way

27 it calculated the billing-month total weather load (as described in Step 1 above).

28 SPS performed the calculation by substituting the calendar-month sales weather

29 normalization regression coefficient (defined in terms of calendar-month heating


Page 46

RD 1 - 367 of 613

4355

1 degree days, cooling degree days, and number of customers) and the calendar-

2 month normal heating degree days and cooling degree days.

3 Q. How did SPS calculate the calendar-month total sales?

4 A. SPS calculated the calendar-month total sales for the customer classes listed

5 earlier by combining the calendar-month base load and calendar-month total

6 weather load components. For the Area Lighting, Primary General Service, and

7 Large General Service — Transmission classes, SPS calculated the forecasted

8 calendar-month sales based on the projected billing-month sales in the same

9 manner as detailed above. However, for these classes, there are no total weather

10 load sales. The calendar-month total sales for these classes were calculated only

11 in terms of their base load, where the billing-month base load equaled the

12 projected billing-month sales.

13 The Street and Area Lighting class is billed on a calendar-month basis in

14 the succeeding month. Therefore, for this class, the calendar-month sales equal

15 the billing-month sales in the succeeding month.

16 Q. Please describe how the other portions of SPS's customer count and sales

17 forecasts are developed.

18 A. The sales forecasts for SPS's firm partial requirements and non-firm wholesale

19 loads are based on contract terms and analysis of historical trends.

20 Q. How are the Texas voltage level MWh sales estimates derived?

21 A. Texas retail sales by customer class are allocated to voltage levels using historical

22 sales proportions. After developing both the SPS system and Texas retail sales

23 estimates by voltage level, SPS derives the MWh sales estimates at the source by


Page 47

RD 1 - 368 of 613

4356

1 applying the voltage level loss factors to the sales estimates. A loss factor

2 represents transmission and distribution losses, plus all unaccounted for energy.

3 Q. How is SPS's system peak demand forecast developed?

4 A. SPS develops the retail peak demand forecast using an econometric model, with

5 monthly historical systern retail peak demand (MW) as the dependent variable,

6 and system retail sales, weather concepts, a linear trend, seasonal binary, and

7 month specific binary variables as explanatory variables. For full requirements

8 wholesale peak demand forecasts at the delivery point, SPS uses historical

9 monthly load factors to develop monthly peak demands based on the projected

10 monthly sales. A load factor is the ratio of sales to the peak demand sustained

11 over a period of time. The formula to calculate the load factors is:

Sales (MWh)

12 Load Factor (%)—

13 Peaks at the delivery point are then grossed up to the source by applying loss

14 factors.

15 The projected load factors are based on historical load factors. The

16 monthly load factors and loss factors used for the forecast period are assumed to

17 be the same as the historical load factors and loss factors.

18 Q. Have you provided SPS's forecasted monthly sales and system peak

19 demands?

20 A. Yes. SPS's sales forecast and system peak demands are provided in Schedule

21 0-7.1.

22 Q. Does this conclude your pre-filed direct testimony?

23 A. Yes.


Page 48

RD 1 - 369 of 613

4357

Peak Demand (MW) x hours per month

A.At-ca‘z kZ(A-- ELL E. MARKS

AFFIDAVIT

STATE OF COLORADO

COUNTY OF DENVER

JANNELL E. MARKS, first being sworn on her oath, states:

I am the witness identified in the preceding testimony. I have read the testimony and the accompanying attachment(s) and am familiar with the contents. Based upon my personal knowledge, the facts stated in the testimony are true. In addition, in my judgment and based upon my professional experience, the opinions and conclusions stated in the testimony are true, valid, and accurate.

Subscribed and sworn to before me this 7.:4 day of August, 2017 by JANNELL E. MARKS.

fLUJJ4VV, UCL,-.1-;) La-

Notary Public, State of Colorado

My Commission Expires:

94 Nr,f;ARYPUE‘ C

STA7E(.7FOOCORADO NOTARY it) 200t4002K4

VYCOSK ,AARES FEBRUARY 2.1. 2021

Marks Direct - Rate Design RD 1 - 370 of 613

Page 49 4358

19J0

IL

E -

1


Weather Normalization of Test Year Sales

Texas Panhandle Weather Data (Amarillo TX and Lubbock TX) Normal Weather Based on a 10-year Historical Average

Month Year

Weather Act Cal HDD65

Weather Act Cal CDD65

Weather Act Cal Precip

Weather Norm Cal HDD65

Weather Norm Cal CDD65

Weather Norm Cal

Precip

Dev HDD65

Dev CDD65

Dev Precip

Apr 2016 195 13 2.74 215 38 1.20 -20 -25 1.54 May 2016 109 99 2.13 74 152 3.02 35 -53 -0.89 Jun 2016 1 391 1.29 0 390 2.43 1 1 -1.14 Jul 2016 0 582 2.76 0 450 3.14 0 132 -0.38

Aug 2016 0 372 3.83 0 441 2.76 0 -69 1.07 Sep 2016 5 225 0.99 19 205 2.18 -14 20 -1.20 Oct 2016 67 108 0.36 191 42 1.58 -124 66 -1.22 Nov 2016 363 4 0.96 483 1 0.64 -120 3 0.32 Dec 2016 794 0 0.33 800 0 0.79 -6 0 -0.46 Jan 2017 748 0 2.88 813 0 0.49 -65 0 2.39 Feb 2017 447 6 0.61 679 0 0.70 -232 6 -0.10 Mar 2017 305 16 1.65 424 8 1.08 -119 8 0.57 Apr 2017 207 44 1.33 215 38 1.20 -8 6 0.13 May 2017 71 96 1.00 74 152 3.02 -3 -56 -2.01 Jun 2017 0 395 1.57 0 390 2.43 0 5 -0.86

Annual (Apr16 - Mar17) 3,034 1,816 20.52 3,698 1,728 20.02 -664 88 0.50 Annual Dev % -17.9% 5.1% 2.5%

!Annual (Ju116 - Jun17) l 3,007 1,848 18.27 3,698 1,728 20.02 -691 120 -1.75 Annual Dev % -18.7% 7.0% -8.7%

EI9

J0 ZL

E -

I



Southwestern Public Service Company Electric Sales - Texas Calendar Month - MONTHLY MWH

Actual Sales Cal Mth

Res


Small C&I


Large C&I


Street

Actual Sales Cal Mtb

Other Sale Auth

Actual Sales Cal Mth Retail

WN Sales Cat Mth

Res

WN Sales Cal Mth

Small C&1

WN Sales Cal Mth

Large C&I

WN Sales Cal Mth

Street

WN Sales Cal Mth

Other Sale Auth

WN Sales Cal Mth Retail

Apr-I6 101,849 229,140 642,922 2,829 29,259 1,005,998 103,131 235,760 643,223 2,829 30,990 1,015,933 May-16 154,630 249,718 671,625 2,829 29,502 1,108,303 163,866 252,502 671,456 2,829 30,054 1,120,706 Jun-16 247,815 268,925 648,811 2,830 30,825 1,199,207 247,662 267,185 648,586 2,830 30,420 1,196,683 Jul-16 313,869 345,041 688,129 2,829 38,267 1,388,135 275,553 323,053 687,288 2,829 36,520 1,325,243 Aug-16 249,577 346,655 718,218 2,828 40,517 1,357,795 269,436 360,180 718,790 2,828 41,961 1,393,196 Sep-16 199,169 281,483 672,532 2,829 34,328 1,190,341 194,296 277,825 672,187 2,829 33,547 1,180,684 Oct-16 147,345 222,651 690,214 2,829 32,154 1,095,193 141,939 219,125 690,209 2,829 30,547 1,084,649 Nov-16 148,662 239,641 672,676 2,828 26,936 1,090,743 154,841 242,402 672,680 2,828 26,936 1,099,687 Dec-16 228,907 270,757 670,339 2,829 31,021 1,203,854 229,400 270,930 670,339 2,829 31,021 1,204,519 Jan-17 228,262 229,360 679,279 2,829 28,169 1,167,899 233,833 230,506 679,280 2,829 28,169 1,174,618 Feb-17 172,567 212,387 579,251 2,831 23,473 990,509 188,686 212,983 579,251 2,831 23,473 1,007,224 Mar-17 145,298 253,572 669,573 2,828 22,088 1,093,357 150,425 258,132 669,573 2,828 22,424 1,103,380 Apr-17 133,571 234,009 623,659 2,828 30,261 1,024,330 133,383 234,549 623,685 2,828 30,310 1,024,756 May-17 163,381 257,615 667,863 2,830 26,581 1,118,269 173,136 258,899 667,476 2,830 26,621 1,128,961 Jun-17 245,326 291,441 695,152 2,828 32,716 1,267,464 244,099 289,547 694,961 2,828 32,355 1,263,789

Test Year Total 2,337,949 3,149,329 8,003,569 33,948 366,538 13,891,333 2,353,070 3,150,582 8,002,862 33,948 366,059 13,906,522 Updated Test Year Total 2,375,934 3,184,612 8,026,886 33,947 366,510 13,987,888 2,389,027 3,178,129 8,025,719 33,947 363,882 13,990,704



EI9

J0 E

LE

- 1 CIII

(9 93)) (12,40)) 2,524 62,892 115,-101)

9,657 (0,543 (8 044) 865,9)

to 7181 116,715) (10,023)

(426)

(10,6)2) 3,675

Apr-16 May-16 Jun-16 Jul-16 Aug-I6 Sep-I 6 Oct-16 Nov-16 Dec-I6 Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17

Test Year Total Updated Test Year Total

Southwestern Public Service Company Electric Sales - Texas Calendar Month - MONTHLY MWH

11,283)

(6,621)

(300)

(1,7)11

(9,276)

(2,704)

168 (cso

153

1,740

225

406

38,316

21,988

841

1,747

(19,85•(t 1 52,) c 72)

(1,444)

4,872

3,658

345

781

5,406

3,526

5

1,607

(6,17111)

(2,761)

(4)

0

(490

(172)

(0)

0

(5,i71)

(1,140

(1 )

0

(16 1l9) ism)) 0

0

(c,127)

(4,c60)

0 (316)

188

(540, (26) (491

(0,7.5)

(1,2(4)

387

1401

1,228

1,894

192

361

(15,121)

11,254)

707

479

(13,094)

6,482

1,167

2,628

-I 2% -2 8% 0 0% 0 0% -5 6%

-5 6% -I I% 0 0% 0 0% -1 8%

0 1% 0 7% 0 0% 0 0% I 3%

13 9% 6 8% 0 1% 0 0% 4 8%

-7 4% -3 8% -0 I% 0 0% -3 4%

2 5% I 3% 0 I% 0 0% 2 3%

3 8% I 6% 0 0% 0 0% 5 3%

-4 0% -I 1% 0 0% 0 0% 0 0%

-0 2% -0 1% 0 0% 0 0% 0 0%

-2 4% -0 5% 0 0% 0 0% 0 0%

-8 5% -0 3% 0 0% 0 0% 0 0%

-3 4% -I 8% 0 0% 0 0% -1 5%

0 1% -0 2% 0 0% 0 0% -0 2%

-5 6% -0 5% 0 1% 0 0% -0 2%

0 5% 0 7% 0 0% 0 0% 1 I%

-0.6% 0.0% 0.0% 0.0% 0.1%

-0.5% 0.2% 0.0% 0.0% 0.7% 115.1881 (2,817)

-0.1%

-0.02%

Act var Act var Act var Act var Act var Act var

fr WN fr WN fr WN Cal Mth Cal Mth fr WN % var % var % var % var % var % var

Res Small C&1 Large C&1 Street Other Sale Auth Retail Res Small C&I Large C&I Street Other Sale Auth Retail

-1 0% -I I% 0 2% 4 7% -2 5% 0 8% 1 0% -0 8% -0 I% -0 6% -1 7% -0 9% 0 00/0 -0 9% 0 3%

EI9

J017L

E-

1



Texas Retail Weather Adjustment Summary

Heating Degree Days 10-yr normal Heating Degree Days Actual Variance from Normal

Apr-16 215 195 (20)

May-16 74

109 35

Jun-16 0 1 1

Jul-16 0 0 0

Aug-16 0 0 0

Sep-16 19 5

(14)

Oct-16 191 67

(124)

Nov-16 483 363 (120)

Dec-16 800 794 (6)

Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-I6 Dec-16 Cooling Degree Days 10-yr normal 38 152 390 450 441 205 42 1 0 Cooling Degree Days Actual 13 99 391 582 372 225 108 4 0 Variance from Normal (25) (53) 1 132 (69) 20 66 3 0

Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16 Precipitation 10-yr normal 1 20 3 02 2 43 3 14 2.76 2 18 1 58 0 64 0 79 Precipitation Actual 2 74 2 13 1 29 2 76 3 83 0 99 0 36 0 96 0 33 Variance from Normal 1 54 (0 89) (1 14) (0 38) 1 07 (1 20) (1 22) 0 32 (0 46)

Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16 Texas Residential Service (1) (1,044) (7,688) 122 30,382 (15,847) 3,929 5,256 (2,494) (236) Texas Residential Space Heat (2) (239) (1,548) 31 7,934 (4,012) 943 149 (3,685) (256) Texas Small General Service (3) (1) (532) 11 2,896 (1,621) 386 472 (353) (33) Texas Secondary General Service (4) & (5) (6,619) (2,257) 1,729 19,122 (11,921) 3,276 3,058 (2,412) (140) CRM%VA weather Adjustment (6) (300) 174 225 811 (556) 341 0 0 0 Small Municipal and School Service (7) (84) (24) 17 79 (67) 34 67 0 0 Large Municipal and School Service (8) (910) (284) 197 950 (798) 387 692 0 0 Large School Service (9) (737) (243) 192 718 (579) 360 849 0 0 Texas Retail Weather Adjustment Total (9,935) (12,403) 2,524 62,892 (35,401) 9,657 10,543 (8,944) (665)

EI9

J0 S

LE

- 1



Texas Retail Weather Adjustment Summary


Jan-17 813 748 (65)

Feb-17 679 447 (232)

Mar-17 424 305 (119)

Updated

Apr-17 215 207

(8)

May-17 74 71

(3)

Jun-17 0 0 (0)

Test Year Total 3,698 3,034 (664)

Test Year Total 3,698 3,007 (691)

Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Total Total Cooling Degree Days 10-yr normal 0 0 8 38 152 390 1,728 1,728 Cooling Degree Days Actual 0 6 16 44 96 395 1,816 1,848 Variance from Normal 0 6 8 6 (56) 5 88 120

Jan-17 Feb-17 Mar-17 Total Total Precipitation 10-yr normal 0 49 0 70 1 08 1 20 3 02 2 43 20 02 20 02 Precipitation Actual 2 88 0 61 1 65 1 33 1 00 1 57 20 52 18 27 Variance from Normal 2 39 (0 10) 0 57 0 13 (2 01) (0 86) 0 50 (1 75)

Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Total Total Texas Residential Service (1) (2,592) (6,806) (1,566) 259 (8,402) 1,014 1,415 2,897 Texas Residential Space Heat (2) (2,979) (9,313) (3,561) (71) (1,353) 213 (16,535) (15,991) Texas Small General Service (3) (352) (1,048) (267) 0 (551) 87 (444) (385) Texas Secondary General Service (4) & (5) (795) 453 (4,293) (540) (739) 1,808 (799) 6,879 CRMWA Weather Adjustment (6) 0 0 0 (26) 393 191 696 1,155 Small Municipal and School Service (7) 0 0 (28) (3) (2) 20 (6) 100 Large Municipal and School Service (8) 0 0 (190) (25) (20) 182 44 1,178 Large School Service (9) 0 0 (119) (21) (18) 159 441 1,350 Texas Retail Weather Adjustment Total (6,718) (16,715) (10,023) (426) (10,692) 3,675 (15,188) (2,817)

ET

U( ) 9L

E -

1 G

II



Texas Residential Service

Heating Degree Days Weather Adjustment (10-yr normal) Apr-16 May-16 Jun-I6 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16

Vanance from Nonnal (20) 35 1 0 0 (14) (124) (120) (6) Heating Deigee Days Weather Coefficients 0 0000000 0 0000000 0 0000000 0.0000000 0 0000000 0 0000000 0 0000016 0 0001264 0 0002506 Res Service Customers 159,330 160,441 161,565 162,349 163,455 163,916 164,453 164,717 165,224 Heating Degree Days Weather Adjustinent (MWh) 0 0 0 0 0 0 -32 -2,494 -236

Cooling Degree Days Weather Adjustment (10-yr normal) Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16

Variance from Normal (25) (53) 1 132 (69) 20 66 3 0 Cooling Degree Days Weather Coefficients 0.0002593 0 0009111 0.0013211 0 0014161 0 0013957 0 0012228 0 0004876 0 0000000 0.0000000 Res Service Customers 159,330 160,441 161,565 162,349 163,455 163,916 164,453 164,717 165,224 Cooling Degree Days Weather Adjustment (MWh) (1,044) (7,688) 122 30,382 (15,847) 3,929 5,288 0 0

TX Residential Service Weather Adjustment (1) (1,044) (7,688) 122 30,382 (15,847) 3,929 5,256 (2,494) (236)

EI9

J0 LI

E- 1 (D

I



Texas Residential Service

Heating Degree Days Weather Adjustment (10-yr normal) Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17

Variance from Nonnal (65) (232) (119) (8) (3) (0) Heating Degree Days Weather Coefficients 0 0002394 0 0001765 0 0000792 0 0000000 0 0000000 0 0000000 Res Service Customers 165,827 166,199 166,880 167,213 167,821 168,285 Heating Degree Days Weather Adjustment (MWh) -2,592 -6,806 -1,569 0 0 0

Cooling Degree Days Weather Adjustment (10-yr normal) Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17

Variance from Normal 0 6 8 6 (56) 5 Cooling Degree Days Weather Coefficients 0 0000000 0 0000000 0 0000021 0.0002709 0 0009006 0 0013188 Res Service Customers 165,827 166,199 166,880 167,213 167,821 168,285 Cooling Degree Days Weather Adjustment (MWh) 0 0 3 259 (8,402) 1,014

TX Residential Service Weather Adjustment (1) (2,592) (6,806) (1.566) 259 (8,402) 1,014

EI9

J0 S

LE

- I C

M



Texas Residential Space Heat

Heating Degree Days Weather Adjustment (10-yr normal) Apr-16 May-16 Jun-I6 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16

Vanance from Normal (20) 35 1 0 0 (14) (124) (120) (6) Heating Degree Days Weather Coefficients 0 0002839 0 0000000 0 0000000 0 0000000 0 0000000 0 0000006 0 0002294 0 0008271 0 0012210 Res Space Heat Customers 42,122 41,336 40,366 39,663 38,866 38,202 37,697 37,205 36,804 Heating Degree Days Weather Adjustment (MWh) -234 0 0 0 0 0 -1,072 -3,685 -256


Vanance from Normal (25) (53) 1 132 (69) 20 66 3 0 Cooling Degree Days Weather Coefficients 0 0000039 0 0007120 0 0013557 0 0015136 0 0014860 0 0012594 0 0004912 0 0000000 0 0000000 Res Space Heat Customers 42,122 41,336 40,366 39,663 38,866 38,202 37,697 37,205 36,804 Cooling Degree Days Weather Adjustment (MWh) (4) (1,548) 31 7,934 (4,012) 943 1,221 0 0

TX Residential Space Heat Weather Adjustment (2) (239) (1,548) 31 7,934 (4,012) 943 149 (3,685) (256)

EI9

J O 6L

E -

I



Texas Residential Space Heat


Variance from Nonnal (65) (232) (119) (8) (3) (0) Heating Degree Days Weather Coefficients 0 0012515 0 0011098 0 0008388 0 0002648 0 0000000 0 0000000 Res Space Heat Customers 36448 36,165 35,737 35,363 35,003 34,527 Heating Degree Days Weather Adjustment (MWh) -2,979 -9,313 -3,561 -71 0 0


Variance from Normal 0 6 8 6 (56) 5 Cooling Degree Days Weather Coefficients 0 0000000 0 0000000 0.0000000 0 0000000 0 0006952 0.0013524 Res Space Heat Customers 36,448 36,165 35,737 35,363 35,003 34,527 Cooling Degree Days Weather Adjustment (MWh) 0 0 0 0 (1,353) 213

TX Residential Space Heat Weather Adjustment (2) (2,979) (9,313) (3,561) (71) (1,353) 213

E19

1 0 0

8E-

I C

M



Texas Small General Service

Heating Degree Days Weather Adjustment (10-yr normal) Apr-16 May-I6 Jun-16 Jul-16 Aug-I6 Sep-16 Oct-16 Nov-16 Dec-16

Variance from Normal (20) 35 1 0 0 (14) (124) (120) (6) Heating Degree Days Weather Coefficients 0 0000000 0 0000000 0 0000000 0 0000000 0 0000000 0 0000000 0 0000011 0.0000907 0 0001775 Small General Service Customers 32,526 32,547 32,530 32,577 32,593 32,527 32,572 32,523 32,540 Heating Degree Days Weather Adjustment (MWh) 0 0 0 0 0 0 -5 -353 -33


Variance from Normal (25) (53) 1 132 (69) 20 66 3 0 Cooling Degree Days Weather Coefficients 0.0000017 0.0003110 0 0005839 0 0006727 0 0007161 0 0006053 0.0002219 0.0000000 0 0000000 Small General Service Customers 32,526 32,547 32,530 32,577 32,593 32,527 32,572 32,523 32,540 Cooling Degree Days Weather Adjustment (MWh) (1) (532) 11 2,896 (1,621) 386 477 0 0

TX Small General Service Weather Adjustment (I) (532) 11 2,896 (1,621) 386 472 (353) (33)

EI9

J0 I

g£ -

1 all



Texas Small General Service

lleating Degree Days Weather Adjustment (10-yr normal) Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17

Variance from Nonnal (65) (232) (119) (8) (3) (0) Heating Degree Days Weather Coefficients 0.0001656 0 0001385 0 0000688 0 0000000 0 0000000 0 0000000 Small General Service Customers 32,600 32,616 32,644 32,646 32,635 32,671 Heating Degree Days Weather Adjustment (MWh) -352 -1,048 -267 0 0 0


Variance from Normal 0 6 8 6 (56) 5 Cooling Degree Days Weather Coefficients 0.0000000 0 0000000 0 0000000 0.0000000 0.0003036 0 0005827 Small General Service Customers 32,600 32,616 32,644 32,646 32,635 32,671 Cooling Degree Days Weather Adjustment (MWh) 0 0 0 0 (551) 87

TX Small General Service Weather Adjustment (352) (1,048) (267) 0 (551) 87

EI9

JO Z

8E - 1 CM



Texas Small Secondary General Service

Heating Degree Days Weather Adjustment (10-yr normal) Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16

Vanance from Normal (20) 35 1 0 0 (14) (124) (120) (6) Heating Degree Days Weather Coefficients 0 0000000 0 0000000 0 0000000 0 0000000 0 0000000 0 0000000 0 0000211 0 0017003 0 0020640 Small Secondary General Customers 11,804 11,841 11,851 11,831 11,843 11,832 11,845 11,829 11,843 Heating Degree Days Weather Adjustment (MWh) 0 0 0 0 0 0 -31 -2,408 -139

Cooling Degree Days Weather Adjustment (10-yr normal) Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-I6 Oct-16 Nov-16 Dec-16

Vanance from Normal (25) (53) 1 132 (69) 20 66 3 0 Cooling Degree Days Weather Coefficients 0 0000310 0 0057000 0 0102184 0.0118566 0 0125795 0 0105503 0 0039489 0.0000000 0 0000000 Small Secondary General Customers 11,804 11,841 11,851 11,831 11,843 11,832 11,845 11,829 11,843 Cooling Degree Days Weather Adjustment (MWh) -9 -3,550 69 18,537 -10,349 2,447 3,085 0 0

Precipitation Weather Adjustment (10-yr normal) Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-I6 Oct-16 Nov-16 Dec-16

Vanance from Normal 1 54 (0 89) (1 14) (0 38) 1 07 (1 20) (1 22) 0.32 (0 46) Precipitation Weather Coefficients -0 364244 -0.123041 -0 122938 -0 123145 -0 122832 -0 058250 0 000000 0 000000 0 000000 Small Secondary General Customers 11,804 11,841 11,851 11,831 11,843 11,832 11,845 11,829 11,843 Precipitation Weather Adjustment (MWh) (6,610) 1,298 1,660 555 (1,555) 825 0 0 0

TX Small Secondary General Service Weather Adjustment (4) (6,619) (2,252) 1,729 19,092 (11,904) 3,272 3,054 (2,408) (139)

Co.)

EI9

JO E

8E -

1 C

III



Texas Small Secondary General Service


Variance from Normal (65) (232) (119) (8) (3) (0) Heating Degree Days Weather Coefficients 0 0007444 0 0000000 0 0000000 0 0000000 0.0000000 0 0000000 Small Secondary General Customers 11,822 11,816 11,853 11,861 11,871 11,877 Heating Degree Days Weather Adjustment (MWh) -575 0 0 0 0 0


Variance from Normal 0 6 8 6 (56) 5 Cooling Degree Days Weather Coefficients 0.0000000 0 0000000 0 0000000 0 0000000 0 0055513 0 0101848 Small Secondary General Customers 11,822 11,816 11,853 11,861 11,871 11,877 Cooling Degree Days Weather Adjustment (MWh) 0 0 0 0 -3,664 553

Precipitation Weather Adjustment (10-yr normal) Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17

Variance from Nonnal 2 39 (0 10) 0 57 0.13 (2 01) (0 86) Precipitation Weather Coefficients -0 007737 -0 394185 -0 640934 -0 346349 -0 122731 -0 122669 Small Secondary General Custorners 11,822 11,816 11,853 11,861 11,871 11,877 Precipitation Weather Adjustment (MWh) (219) 453 (4,293) (540) 2,931 1,255

TX Small Secondary General Service Weather Adjustment (4) (794) 453 (4,293) (540) (733) 1,807

£I9

JO 178

E - 1

CM



Texas Large Secondary General Service

Heating Degree Days Weather Adjustment (10-yr normal) Apr-I6 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16

Vanance from Nonnal (20) 35 1 0 0 (14) (124) (120) (6) Heating Degree Days Weather Coefficients 0 0000000 0 0000000 0 0000000 0 0000000 0 0000000 0 0000000 0 0000211 0.0017003 0 0020640 Large Secondary General Customers 19 19 19 19 19 18 18 18 18 Heating Degree Days Weather Adjustment (MWh) 0 0 0 0 0 0 0 -4 0


Variance from Normal (25) (53) 1 132 (69) 20 66 3 0 Cooling Degree Days Weather Coefficients 0 0000310 0.0057000 0 0102184 0 0118566 0.0125795 0 0105503 0.0039489 0 0000000 0 0000000 Large Secondary General Customers 19 19 19 19 19 18 18 18 18 Cooling Degree Days Weather Adjustment (MWh) 0 -6 0 30 -17 4 5 0 0

Precipitation Weather Adjustment (10-yr normal) Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16

Variance from Normal Precipitation Weather Coefficients

Large SG customer usage is not nnpacted by precipitation Small Secondary General Customers Precipitation Weather Adjustment (MWh)

TX Large Secondary General Service Weather Adjustment (5) (0) (6) 0 30 (17) 4 5 (4) (0)

£19

J0 g

gE

- 1

CM



Texas Large Secondary General Service


Variance from Normal (65) (232) (119) (8) (3) (0) Heating Degree Days Weather Coefficients 0 0007444 0 0000000 0 0000000 0 0000000 0 0000000 0 0000000 Large Secondary General Custorners 18 19 19 19 19 19 Heating Degree Days Weather Adjustment (MWh) -1 0 0 0 0 0


Variance from Normal 0 6 8 6 (56) 5 Cooling Degree Days Weather Coefficients 0.0000000 0 0000000 0 0000000 0 0000000 0 0055513 0 0101848 Large Secondary General Customers 18 19 19 19 19 19 Cooling Degree Days Weather Adjustment (MW11) 0 0 0 0 -6 1


Variance from Normal Precipitation Weather Coefficients Small Secondary General Customers

Large SG customer usage is not nnpacted by precipitation

Precipitation Weather Adjustment (MWh)

TX Large Secondary General Service Weather Adjustment (5) (1 ) 0 0 0 (6) 1

EI9

J0 98E

- 1 C

RT



Canadian River Municipal Water Authority (CRMWA)

Cooling Degree Days Weather Adjustment (10-yr normal)

Variance from Normal Cooling Degree Days Weather Coefficients

Apr-16 (25)

0.0000000

May-16 (53)

0 0000000

Jun-16 1

4 9067682

Jul-16 132

5 5775218

Aug-16 (69)

4 9949963

Sep-16 20

5 4812912

Oct-16 66

0 0000000

Nov-16 3

0 0000000

Dec-16 0

0 0000000 CRMWA Customers 1 1 1 1 1 1 1 I 1

ooling Degree Days Weather Adjustment (MWh) 0 0 3 737 (347) 107 0 0 0

Precipitation Weather Adjustment (10-yr normal) Apr-I6 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16

Variance from Normal 1 54 (0 89) (1 14) (0 38) 1 07 (1 20) (1 22) 0 32 (0 46) Precipitation Weather Coefficients -195 32244 -195 32244 -195 32244 -195 32244 -195 32244 -195 32244 0 00000 0 00000 0 00000 Small Secondary General Customers 1 I 1 1 1 1 1 1 1 Precipitation Weather Adjustment (MWh) (300) 174 223 74 (209) 234 0 0 0

CRMWA Weather Adjustment (6) (300) 174 225 811 (556) 341 0 0 0

EI9

JO L

8E -

1



Canadian River Municipal Water Authority (CRMWA)


Variance from Normal 0 6 8 6 (56) 5 Cooling Degree Days Weather Coefficients 0 0000000 0 0000000 0 0000000 0 0000000 0 0000000 4 9067682 CRMWA Customers I 1 1 1 1 1 Cooling Degree Days Weather Adjustment (MWh) 0 0 0 0 0 22


Variance from Normal 2 39 (0 10) 0 57 0 13 (2 01) (0 86) Precipitation Weather Coefficients 0.00000 0 00000 0 00000 -195 32244 -195 32244 -195 32244 Small Secondary General Customers 1 1 1 1 1 1 Precipitation Weather Adjustment (MWh) 0 0 0 (26) 393 168

CRMWA Weather Adjustment (6) 0 0 0 (26) 393 191

I -(1

21-1

AIH

f lu

mtP

LIW

Precipitation Weather Adjustment (10-yr normal) Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16

Variance from Normal 1 54 (0 89) (1 14) (0 38) 1 07 (1 20) (1 22) 0 32 (0 46) Precipitation Weather Coefficients (0.20546) (0 11023) (0 07677) (0.07677) (0 03620) 0.00000 0 00000 0.00000 0 00000 Small Secondary General Customers 4,537 4,537 4,544 4,545 4,541 4,540 4,528 4,529 4,523 Precipitation Weather Adjustment (MWh) (1,433) 446 397 133 (176) 0 0 0 0

Weather Adjustment (1,731) (551) 406 1,747 (1,444) 781 1,607 0 0

Texas - Municipal and School Weather Impact Allocation (Billed Sales in KWh) Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16

SMSTX LMSTX LSSTX Total

1,317 14,219 11,523 27,059

1,307 15,308 13,088 29,703

1,282 14,752 14,379 30,413

1,475 17,834 13,481 32,790

1,733 20,555 14,922 37,209

1,646 18,567 17,304 37,517

1,393 14,352 17,615 33,361

1,267 14,028 14,543 29,838

1,453 12,717 13,499 27,668

Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16 SMSTX 4 87% 4 40% 4 22% 4 50% 4 66% 4 39% 4 18% 4 24% 5 25% LMSTX 52 55% 51 54% 48 51% 54 39% 55 24% 49 49% 43 02% 47.02% 45 96% LSSTX 42 58% 44 06% 47 28% 41 11% 40 10% 46 12% 52 80% 48 74% 48 79% Total 100 00% 100 00% 100 00% 100 00% 100.00% 100 00% 100.00% 100 00% 100.00%

Municipal and School Allocation of Sales impacted by weather Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-I6

SMSTX (7) (84) (24) 17 79 (67) 34 67 0 0 LMSTX (8) (910) (284) 197 950 (798) 387 692 0 0 LSSTX (9) (737) (243) 192 718 (579) 360 849 0 0

£I9

J0 8

8£ - I C

M



Texas Municipal and School Service


Vanance from Nonnal ooling Degree Days Weather Coefficients

Municipal & School Customers ooling Degree Days Weather Adjustment (MWh)

Apr-16 (25)

May-16 (53)

Jun-16 1

Jul-16 132

Aug-16 (69)

Sep-16 20

Oct-16 66

Nov-16 3

Dec-16 0

0 0025980 0.0041771 0 0031655 0 0026868 0 0040204 0 0087790 0 0053828 0 0000000 0 0000000 4,537 4,537 4,544 4,545 4,541 4,540 4,528 4,529 4,523 (298) (997) 8 1,614 (1,268) 781 1,607 0 0

£I9

Jo

68

£ - I

CM

Southwestern Public

Weather Normalization

Service Company

of Test Year Sales

Texas Municipal and School Service


Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Variance from Normal 0 6 8 6 (56) 5 Cooling Degree Days Weather Coefficients 0 0000000 0 0000000 0 0000217 0 0027429 0 0041896 0 0031806 Municipal & School Customers 4,520 4,515 4,477 4,482 4,478 4,481 Cooling Degree Days Weather Adjustment (MWh) 0 0 1 70 (1,043) 65


Variance from Normal 2 39 (0 10) 0 57 0 13 (2 01) (0 86) Precipitation Weather Coefficients 0,00000 0 00000 (0 13317) (0 20247) (0.11135) (0 07677) Srnall Secondary General Customers 4,520 4,515 4,477 4,482 4,478 4,481 Precipitation Weather Adjustment (MWh) 0 0 (337) (119) 1,003 296

Weather Adjustment 0 0 (336) (49) (40) 361

Texas - Municipal and School Weather Impact Allocation (Billed Sales Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17

SMSTX 2,090 1,979 1,808 1,521 1,683 1,604 LMSTX 14,620 11,914 12,398 13,986 14,021 14,681 LSSTX 15,843 13,709 7,746 11,318 13,032 12,850 Total 32,553 27,603 21,952 26,825 28,735 29,135

Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 SMSTX 6.42% 7 17% 8 24% 5 67% 5 86% 5 51% LMSTX 44 91% 43 16% 56 48% 52 14% 48 79% 50.39% LSSTX 48 67% 49 67% 35 29% 42 19% 45 35% 44 11% Total 100.00% 100.00% 100 00% 100 00% 100.00% 100 00%

Municipal and School Allocation of Sales impacted by weather Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17

SMSTX (7) 0 0 (28) (3) (2) 20 LMSTX (8) 0 0 (190) (25) (20) 182 LSSTX (9) 0 0 (119) (21) (18) 159

£I9JO

06E

- 1 (D

I



Year Month Date

Texas Residential Service Coefficients (Modeled with MWH)

Model Coefficients Monthly weights Model to calendar Calendar Month HDD CDD Current 1st Future 2nd Future HDD CDD

2016 4 Apr-16 0.000000 0 000000 44.76% 54 92% 0 32% 0 0000000 0 0002593 2016 5 May-16 0 000000 0.000465 41 47% 58 53% 0 00% 0 0000000 0 0009111 2016 6 Jun-16 0 000000 0.001227 46 67% 52.86% 0.48% 0 0000000 0 0013211 2016 7 Jul-16 0 000000 0 001403 42 86% 57.14% 0 00% 0 0000000 0.0014161 2016 8 Aug-16 0 000000 0 001426 47.16% 52.69% 0 15% 0 0000000 0.0013957 2016 9 Sep-16 0 000000 0 001369 47 30% 52 54% 0 16% 0 0000000 0 0012228 2016 10 Oct-16 0 000000 0.001095 44.55% 54 69% 0 77% 0 0000016 0.0004876 2016 11 Nov-16 0 000000 0 000000 37 94% 61.59% 0.48% 0 0001264 0.0000000 2016 12 Dec-16 0 000203 0 000000 39 29% 60.71% 0.00% 0 0002506 0.0000000 2017 1 Jan-17 0 000281 0 000000 45.78% 53.00% 1.23% 0 0002394 0 0000000 2017 2 Feb-17 0 000205 0 000000 37.41% 62 59% 0 00% 0 0001765 0 0000000 2017 3 Mar-17 0.000160 0 000000 49.62% 49 92% 0 46% 0.0000792 0 0000021 20 I 7 4 Apr-17 0 000000 0.000000 41 75% 58 25% 0.00% 0 0000000 0.0002709 2017 5 May-17 0 000000 0.000465 42.86% 57.14% 0.00% 0 0000000 0 0009006 2017 6 Jun-17 0 000000 0.001227 47 94% 51.59% 0 48% 0.0000000 0 0013188 2017 7 Jul-17 0.000000 0 001403 44.09% 55.91% 0 00% 2017 8 Aug-17 0 000000 0.001426 48.39% 51.15% 0 46% 2017 9 Sep-17 0 000000 0 001369 44 29% 55.71% 0 00%

Texas Resi Model Coefficients

dential Space Heat Coefficients (Modeled with MWH) Monthly weights Model to calendar Calendar Month



£19J0

16E

- 1 (R

I

Year Month HDD CDD Current 1st Future 2nd Future HDD CDD 2016 4 0 000634 0 000000 44 76% 54 92% 0 32% 0 0002839 0 0000039 2016 5 0 000000 0 000000 41 47% 58 53% 0 00% 0.0000000 0.0007120 2016 6 0 000000 0 001217 46.67% 52 86% 0 48% 0 0000000 0.0013557 2016 7 0 000000 0.001477 42 86% 57 14% 0 00% 0 0000000 0.0015136 2016 8 0.000000 0 001541 47 16% 52 69% 0 15% 0 0000000 0 0014860 2016 9 0 000000 0 001438 47 30% 52 54% 0 16% 0 0000006 0.0012594 2016 10 0 000000 0 001103 44.55% 54 69% 0.77% 0.0002294 0.0004912 2016 11 0 000404 0 000000 37 94% 61 59% 0 48% 0 0008271 0 0000000 2016 12 0 001084 0 000000 39.29% 60 71% 0 00% 0 0012210 0 0000000 2017 1 0 001310 0 000000 45.78% 53 00% 1 23% 0 0012515 0 0000000 2017 2 0 001206 0 000000 37 41% 62 59% 0 00% 0.0011098 0 0000000 2017 3 0 001052 0 000000 49 62% 49 92% 0 46% 0 0008388 0 0000000 2017 4 0 000634 0 000000 41.75% 58 25% 0.00% 0.0002648 0 0000000 2017 5 0 000000 0 000000 42 86% 57 14% 0.00% 0 0000000 0.0006952 2017 6 0.000000 0.001217 47 94% 51 59% 0 48% 0 0000000 0.0013524 2017 7 0 000000 0.001477 44 09% 55 91% 0.00% 2017 8 0.000000 0 001541 48 39% 51 15% 0 46% 2017 9 0 000000 0.001438 44 29% 55.71% 0 00%

EI9

J 0 Z

6E

- 1 M

I



Texas Small General Service Coefficients (Modeled with MWH) Model Coefficients Monthly weights Model to calendar Calendar Month

Year Month HDD CDD Precip Current 1st Future 2nd Future HDD CDD Precip 2016 4 0 000000 0.000000 0 000000 44 76% 54.92% 0.32% 0.0000000 0 0000017 0 0000000 2016 5 0 000000 0.000000 0 000000 41 47% 58 53% 0 00% 0 0000000 0.0003110 0 0000000 2016 6 0.000000 0 000531 0 000000 46.67% 52 86% 0 48% 0 0000000 0.0005839 0 0000000 2016 7 0.000000 0 000629 0 000000 42.86% 57 14% 0 00% 0.0000000 0.0006727 0.0000000 2016 8 0 000000 0 000705 0 000000 47.16% 52.69% 0 15% 0.0000000 0.0007161 0 0000000 2016 9 0.000000 0 000726 0 000000 47.30% 52 54% 0 16% 0.0000000 0.0006053 0 0000000 2016 10 0 000000 0 000498 0 000000 44 55% 54.69% 0 77% 0 0000011 0 0002219 0 0000000 2016 11 0 000000 0 000000 0 000000 37 94% 61 59% 0.48% 0 0000907 0 0000000 0.0000000 2016 12 0 000146 0 000000 0 000000 39 29% 60 71% 0.00% 0 0001775 0 0000000 0.0000000 2017 1 0.000198 0 000000 0.000000 45 78% 53 00% 1 23% 0.0001656 0 0000000 0 0000000 2017 2 0.000138 0 000000 0.000000 37.41% 62 59% 0 00% 0.0001385 0.0000000 0 0000000 2017 3 0 000139 0 000000 0.000000 49 62% 49 92% 0 46% 0 0000688 0 0000000 0 0000000 2017 4 0 000000 0 000000 0.000000 41 75% 58 25% 0 00% 0 0000000 0 0000000 0.0000000 2017 5 0.000000 0.000000 0 000000 42 86% 57 14% 0 00% 0.0000000 0 0003036 0 0000000 2017 6 0.000000 0.000531 0 000000 47 94% 51 59% 0 48% 0.0000000 0 0005827 0 0000000 2017 7 0 000000 0.000629 0 000000 44.09% 55.91% 0.00% 2017 8 0 000000 0.000705 0 000000 48.39% 51 15% 0 46% 2017 9 0 000000 0.000726 0 000000 44 29% 55.71% 0 00%

EI9

J0 E

6E -

I C

M



Year Month HDD Model Coefficient.

Texas Secondary General Service

Monthly wetghts Model to calendar Coefficients

HDD

(Modeled with Calendar Month

MWH)

Preen:, Actual Cust HDD CDD CDD Preen:, Current 1st Future 2nd Future CDD 2016 4 0 000000 0.000000 -7820 015465 44 76% 54 92% 0 32% 0 0000000 0 3666942 -4306 4612103 11,823 0 0000000 0 0000310 2016 5 0 000000 0 000000 -1459 270694 41 47% 58 53% 0 00% 0 0000000 67 6018468 -1459 2706936 11,860 0 0000000 0 0057000 2016 6 0 000000 115 508667 -1459 270694 46 67% 52 86% 0 48% 0 0000000 121 2927689 -1459 2706936 11,870 0 0000000 0 0102184 2016 7 0 000000 126 129304 -1459 270694 42 86% 57 14% 0 00% 0 0000000 140.5007619 -1459 2706936 11,850 0 0000000 0 0118566 2016 8 0.000000 151 279356 -1459.270694 47 16% 52.69% 0 15% 0 0000000 149 2174363 -1457 0291104 11,862 0 0000000 0.0125795 2016 9 0 000000 147.500374 -1459.270694 47 30% 52 54% 0 16% 0 0000000 125 0209773 -690 2582011 11,850 0 0000000 0 0105503 2016 10 0 000000 105 160435 0 000000 44.55% 54 69% 0 77% 0 2500623 46.8456625 0 0000000 11,863 0 0000211 0 0039489 2016 11 0.000000 0 000000 0.000000 37.94% 61 59% 0 48% 20 1433459 0 0000000 0 0000000 11,847 0 0017003 0 0000000 2016 12 32.558111 0 000000 0 000000 39 29% 60.71% 0 00% 24 4811003 0.0000000 0 0000000 11,861 0 0020640 0 0000000 2017 1 19 253593 0 000000 0.000000 45 78% 53.00% 1 23% 8 8134728 0.0000000 -91 6023047 11,840 0 0007444 0 0000000 2017 2 0.000000 0 000000 0 000000 37.41% 62 59% 0 00% 0 0000000 0.0000000 -4665 1745191 11,835 0 0000000 0 0000000 2017 3 0.000000 0 000000 -7454 137547 49 62% 49 92% 0 46% 0.0000000 0 0000000 -7609.1693791 11,872 0 0000000 0 0000000 2017 4 0 000000 0 000000 -7820 015465 41 75% 58 25% 0.00% 0 0000000 0 0000000 -4114 6292251 11,880 0 0000000 0 0000000 2017 5 0 000000 0 000000 -1459 270694 42 86% 57 14% 0 00% 0 0000000 66.0049528 -1459 2706936 11,890 0 0000000 0 0055513 2017 6 0 000000 115.508667 -1459.270694 47 94% 51.59% 0 48% 0 0000000 121 1579037 -1459.2706936 11,896 0.0000000 0 0101848 2017 7 0 000000 126 129304 -1459 270694 44 09% 55 91% 0.00% 2017 8 0 000000 151 279356 -1459 270694 48 39% 51 15% 0 46% 2017 9 0 000000 147 500374 -1459 270694 44 29% 55 71% 0.00%



EI9

J0 1

76£

- I (RI

Precip -0.3642444 -0.1230414 -0 1229377 -0 1231452 -0.1228317 -0.0582496 0 0000000 0,0000000 0 0000000 -0 0077367 -0 3941846 -0 6409341 -0 3463493 -0 1227309 -0 1226690

EI9

J 0 g

6£ - 1 CM



Year Month Model Coefficients (Calendar)

Texas CRMWA Coefficients (Modeled with MWH) Monthly weights Model to calendar Revenue Month

Precip HDD CDD Precip Current 1st Future 2nd Future HDD CDD 2016 4 0.000000 0 000000 -195 322444 44 76% 54.92% 0.32% 0 0000000 0 0000000 -195 3224443 2016 5 0 000000 0 000000 -195 322444 41 47% 58 53% 0 00% 0 0000000 4 9067682 -195 3224443 2016 6 0 000000 4 906768 -195 322444 46 67% 52 86% 0 48% 0 0000000 5 5775218 -195 3224443 2016 7 0 000000 5 577522 -195 322444 42 86% 57 14% 0 00% 0 0000000 4 9949963 -195 3224443 2016 8 0 000000 4 994996 -195 322444 47 16% 52 69% 0 15% 0 0000000 5 4812912 -195 3224443 2016 9 0 000000 5 481291 -195 322444 47 30% 52.54% 0 16% 0.0000000 0 0000000 0 0000000 2016 10 0 000000 0 000000 0 000000 44 55% 54 69% 0 77% 0 0000000 0 0000000 0 0000000 2016 11 0 000000 0.000000 0 000000 37 94% 61 59% 0 48% 0.0000000 0 0000000 0 0000000 2016 12 0 000000 0 000000 0.000000 39 29% 60 71% 0.00% 0.0000000 0 0000000 0 0000000 2017 1 0.000000 0 000000 0 000000 45 78% 53 00% 1 23% 0 0000000 0.0000000 0 0000000 2017 2 0 000000 0 000000 0 000000 37 41% 62 59% 0 00% 0.0000000 0 0000000 0 0000000 2017 3 0 000000 0 000000 0 000000 49 62% 49 92% 0 46% 0 0000000 0 0000000 -195 3224443 2017 4 0 000000 0 000000 -195 322444 41 75% 58 25% 0 00% 0 0000000 0.0000000 -195 3224443 2017 5 0 000000 0 000000 -195 322444 42 86% 57 14% 0 00% 0 0000000 4 9067682 -195 3224443 2017 6 0 000000 4 906768 -195 322444 47 94% 51 59% 0 48% 0 0000000 5.5775218 -195 3224443 2017 7 0 000000 5.577522 -195 322444 44.09% 55 91% 0 00% 2017 8 0 000000 4 994996 -195 322444 48 39% 51 15% 0 46% 2017 9 0.000000 5 481291 -195 322444 44 29% 55.71% 0 00%

EI9

J0 9

6E

- 1 C

[11



Year Month

Texas Mun icipals and Schools Coefficients (Modeled Model Coefficients Monthly weights Model to calendar

with MWH)

HDD Calendar Month

HDD CDD Precip Current 1st Futttre 2nd Ftttttre CDD Precip 2016 4 0 000000 0.000000 -0 265295 44 76% 54 92% 0 32% 0.0000000 0.0025980 -0 2054648 2016 5 0 000000 0 004709 -0 157446 41 47% 58 53% 0 00% 0.0000000 0 0041771 -0 1102318 2016 6 0.000000 0 003800 -0 076773 46.67% 52 86% 0 48% 0 0000000 0 0031655 -0 0767732 2016 7 0 000000 0.002609 -0 076773 42 86% 57.14% 0 00% 0 0000000 0 0026868 -0 0767732 2016 8 0 000000 0 002745 -0.076773 47 16% 52.69% 0.15% 0 0000000 0 0040204 -0 0362049 2016 9 0.000000 0 005138 0 000000 47 30% 52 54% 0 16% 0.0000000 0 0087790 0.0000000 2016 10 0 000000 0 012083 0.000000 44 55% 54 69% 0 77% 0 0000000 0 0053828 0 0000000 2016 11 0 000000 0.000000 0 000000 37.94% 61 59% 0.48% 0 0000000 0 0000000 0 0000000 2016 12 0.000000 0.000000 0 000000 39.29% 60 71% 0.00% 0.0000000 0 0000000 0 0000000 2017 1 0.000000 0.000000 0.000000 45 78% 53 00% 1.23% 0.0000000 0 0000000 0 0000000 2017 2 0.000000 0 000000 0.000000 37 41% 62.59% 0 00% 0.0000000 0.0000000 0.0000000 2017 3 0.000000 0 000000 0.000000 49 62% 49.92% 0.46% 0.0000000 0.0000217 -0 1331691 2017 4 0 000000 0.000000 -0 265295 41 75% 58.25% 0 00% 0 0000000 0 0027429 -0.2024684 2017 5 0 000000 0 004709 -0 157446 42 86% 57 14% 0.00% 0 0000000 0 0041896 -0.1113471 2017 6 0 000000 0 003800 -0 076773 47 94% 51 59% 0.48% 0 0000000 0 0031806 -0 0767732 2017 7 0.000000 0 002609 -0 076773 44 09% 55.91% 0.00% 2017 8 0 000000 0.002745 -0 076773 48 39% 51 15% 0 46% 2017 9 0.000000 0 005138 0 000000 44 29% 55 71% 0 00%

ET9J0 £

6E

- 1

CM


Weather Normalization of Test Year Sales - Wholesale and New Mexico

Roswell Weather Data Normal Weather Based on a 10-year Historical Average

Month Year






Weather Norm Cal

Precip

Dev HDD65

Dev CDD65

Dev Precip

Apr 2016 Apr2016 137 20 0 77 124 58 0 40 13 -38 0 37 May 2016 May2016 62 158 0 60 31 202 1 48 31 -44 -0 88 Jun 2016 Jun2016 0 478 0 53 0 483 1 43 0 -5 -0 90 Jul 2016 Ju12016 0 672 0 45 0 506 1 96 0 166 -1 51

Aug 2016 Aug2016 0 427 5 05 0 501 1 68 0 -74 3 37 Sep 2016 Sep2016 2 231 1 21 7 253 2 03 -5 -22 -0 82 Oct 2016 0ct2016 35 78 0 46 129 53 1.10 -94 25 -0.64 Nov 2016 Nov2016 349 2 0 71 448 0 0 22 -99 2 0 49 Dec 2016 Dec2016 732 0 0 30 749 0 0 49 -17 0 -0 19 Jan 2017 Jan2017 647 0 1 09 749 0 0 28 -102 0 0 81 Feb 2017 Feb2017 380 0 0 16 537 0 0 32 -157 0 -0.16 Mar 2017 Mar2017 175 34 0 32 324 5 0 31 -149 29 0 01 Apr 2017 Apr2017 112 74 1 45 124 58 0 40 -12 16 1 05 May 2017 May2017 10 173 0 54 31 202 1 48 -21 -29 -0 94 Jun 2017 Jun2017 0 505 0 56 0 483 1 43 0 22 -0 87

Annual (Apr16 - Mar17) 2,519 2,100 11.65 3,097 2,061 11.71 -578 39 -0.06 Annual Dev % -18.7% 1.9% -0.5%

lAnnual (Jull6 - Jun17) i 2,442 2,196 12.30 3,097 2,061 11.71 -655 135 0.59 Annual Dev % -21.2% 6.6% 5.1%

CA) 00 U1

£19

.10 8

6£ - I



Lubbock Weather Data Normal Weather Based on a 10- ear Historical Average

Month Year






Weather Norm Cal

Precip

Dev HDD65

Dev CDD65

Dev Precip

Apr 2016 Apr2016 122 35 1 02 153 69 1 21 -31 -34 -0 19 May 2016 May2016 83 176 3 66 50 212 3 47 33 -36 0 19 Jun 2016 Jun2016 3 426 1 04 0 455 1 99 3 -29 -0 95 Jul 2016 Ju12016 0 642 0 58 0 490 2 24 0 152 -1 66

Aug 2016 Aug2016 0 431 3 03 0 496 1 41 0 -65 1 62 Sep 2016 Sep2016 2 254 1 47 11 241 3 04 -9 13 -1.57 Oct 2016 0ct2016 44 127 1 05 144 63 1 62 -100 64 -0.57 Nov 2016 Nov2016 318 7 0 54 420 3 0 53 -102 4 0 01 Dec 2016 Dec2016 706 0 0 49 724 0 0 89 -18 0 -0 40 Jan 2017 Jan2017 653 0 2 03 738 0 0 53 -85 0 1 50 Feb 2017 Feb2017 361 4 0 89 583 0 0 69 -222 4 0 20 Mar 2017 Mar2017 220 46 0 67 347 14 1.24 -127 32 -0 57 Apr 2017 Apr2017 141 79 1.32 153 69 1 21 -12 10 0 11 May 2017 May2017 23 173 0 58 50 212 3 47 -27 -39 -2.89 Jun 2017 Jun2017 0 478 1 78 0 455 1 99 0 23 -0 21

Annual (Apr16 - Mar17) 2,512 2,148 16.47 3,171 2,043 18.87 -659 106 -2.40 Annual Dev % -20.8% 5.2% -12.7%

'Annual (Ju116 - Jun17) l 2,468 2,241 14.43 3,171 2,043 18.87 -703 199 -4.44 Annual Dev % -22.2% 9.7% -23.5%

4=6 CA4 00

EI9

J0 6

6E

- I U

N



Amarillo Weather Data Normal Weather Based on a 10-year Historical Avera e

Month Year

Weather Act Cal HDD6S





Weather Norm Cal

Precip

Dev HDD65

Dev CDD65

Dev Precip

Apr 2016 Apr2016 224 8 3 33 238 30 1 20 -14 -22 2 13 May 2016 May2016 126 77 1 60 85 132 2.86 42 -55 -1 26 Jun 2016 Jun2016 0 376 1 38 I 367 2 59 -1 9 -I 21 Jul 2016 Ju12016 0 561 3 51 0 435 3 45 0 126 0 06

Aug 2016 Aug2016 0 349 4.10 0 422 3.22 0 -73 0.88 Sep 2016 Sep2016 9 212 0 82 22 194 1 89 -13 18 -1 07 Oct 2016 0ct2016 77 105 0 13 210 36 1 57 -133 69 -1 44 Nov 2016 Nov2016 379 3 1 10 506 1 0 68 -127 2 0 42 Dec 2016 Dec2016 826 0 0 27 826 0 0 75 0 0 -0 48 Jan 2017 Jan2017 787 0 3 17 839 0 0 47 -52 0 2 70 Feb 2017 Feb2017 480 5 0 51 714 0 0 71 -234 5 -0 20 Mar 2017 Mar2017 347 12 1 98 451 6 1 03 -104 6 0 95 Apr 2017 Apr2017 238 37 1 34 238 30 1.20 0 7 0 14 May 2017 May2017 94 76 1 15 85 132 2 86 10 -56 -I 71 Jun 2017 Jun2017 1 367 1 50 1 367 2 59 0 0 -1 09

Annual (Apr16 - Mar17) 3,255 1,708 21.90 3,891 1,623 20.41 -636 85 1.49 Annual Dev % -16.4% 5.2% 7.3%

lAnnual (Jul16 - Jun17) l 3,238 1,727 19.58 3,891 1,623 20.41 -653 104 -0.83 Annual Dev % -16.8% 6.4% -4.1%

CA4 00

ET

9J0

OM

? - 1 (III



New Mexico - Calendar Month Weather Normal Sales 10-Year Normal Weather

Southwestern Public Service Company Electric Sales - New Mexico Calendar Month - MONTHLY MWH


Res


Small C&1


Large C&1

Actual Sales Cal Mth Street


Other Sale Auth


Retail

WN Sales Cal Mth

Res

WN Sales Cal Mth

Small C&I

WN Sales Cal Mth

Large C&I

WN Sales Cal Mth Street

WN Sales Cal Mth

Other Sale Auth

WN Sales Cal Mth

Retail Apr-16 54,974 123,731 199,673 1,110 9,472 388,960 54,882 124,085 199,673 1,110 9,753 389,502 May-16 72,248 134,953 213,304 1,108 10,242 431,856 74,102 135,257 213,304 1,108 9,883 433,655 Jun-16 102,682 136,430 209,813 1,111 11,693 461,728 103,049 136,108 209,813 1,111 11,441 461,522 Jul-16 130,729 165,562 215,037 1,111 14,276 526,716 115,639 158,355 215,035 1,111 13,443 503,585

Aug-16 109,793 143,580 226,796 1,110 13,670 494,949 116,307 149,503 226,797 1,110 14,134 507,851 Sep-16 79,173 134,923 210,907 1,110 11,498 437,611 80,755 135,222 210,907 1,110 11,877 439,872

Oct-16 64,760 119,733 213,200 1,110 10,829 409,633 64,728 119,380 213,200 1,110 10,531 408,949 Nov-16 74,309 115,501 207,431 1,109 8,607 406,956 77,789 115,618 207,431 1,109 8,607 410,553 Dec-16 113,833 124,172 211,058 1,107 10,414 460,584 114,818 124,212 211,058 1,107 / 0,414 461,609 Jan-17 111,498 124,679 209,623 1,111 9,465 456,376 117,798 124,990 209,623 1,111 9,474 462,996 Feb-17 87,368 112,821 193,947 1,110 6,772 402,018 95,794 113,034 193,947 1,110 6,685 410,570 Mar-17 54,503 136,955 214,431 1,111 9,882 416,883 59,405 137,177 214,431 1,111 9,887 422,013 Apr-17 59,257 126,009 199,934 1,112 9,991 396,303 59,340 126,982 199,934 1,112 10,782 398,149 May-17 72,413 133,775 215,699 1,107 11,577 434,571 73,607 133,736 215,699 1,107 11,099 435,248 Jun-17 107,144 148,829 229,924 1,112 10,938 497,946 105,360 147,541 229,924 1,112 10,531 494,467

Test Year Total 1,055,870 1,573,040 2,525,218 13,319 126,821 5,294,268 1,075,067 1,572,943 2,525,218 13,319 126,130 5,312,677 Updated Test Year Total 1,064,781 1,586,539 2,547,985 13,320 127,919 5,340,545 1,081,342 1,585,753 2,547,985 13,320 127,464 5,355,863

c)

H

ar'Q' 72, 4 CD •

C.) 5:4, En UN co

EI9

JO

1017

- I



New Mexico - Calendar Month Weather Normal Sales

10-Year Normal Weather

Southwmtern Public Service Company Electric Sales - New Mexico

Calendar Month - MONTHLY MWH

Act var

fr WN Res

Act var fr WN

Small C&I

Act var fr WN

Large C&I

Act var fr WN Street

Act var fr WN

Other Sale Auth

Act var fr WN Retail

% vsr Res

%var Small C&1

% var Large C&I

% var Street

%var Other Sale Auth

% var Retail

Apr-16 92 (354) (0) (280) (542) 0 17% -0 29% 0 00% 0 00% -2 87% -0.14%

May-16 ((,854) (304) (0) 359 (1,709) -2.50% -0 22% 0.00% 0.00% 3.63% -0 41%

Jun-16 (3(7) 322 (0) 252 206 -0.36% 0 24% 0 00% 0 00% 2.20% 0 04%

Jul-16 15,090 7,207 1 833 23,131 13 05% 4.55% 0 00% 0.00% 6 20% 4 59%

Aug-16 0).5(5) (5,023) (1) ((64) (12,)>021 -5 60% -3 96% 0 00% 0.00% -3 28% -2 54%

Sep-16 (1.582) (300) (0) (370) (2,201) -1.96% -0 22% 0 00% 0 00% -3 19% -0.51%

Oct-I6 32 353 0 298 684 0 05% 0 30% 0 00% 0 00% 2 83% 0.17%

Nov-16 (3.480) (1(7) 0 0 (3,5971 -4 47% -0 10% 0 00% 0 00% 0.00% -0 88%

Dec-16 (985) (101 0 0 (1,025) -0 86% -0 03% 0 00% 0 00% 0 00% -0.22%

Jan-17 (0.101) ((ll) 0 (9) (6.620) -5.35% -0 25% 0 00% 0.00% -0 09% -1 43%

Feb-17 (8.426) (214) 0 87 (8.552) -8.80% -0 19% 0 00% 0 00% 1 31% -2 08%

Mar-17 (4.902) (222) 0 (5) (5,130) -8 25% -0 16% 0 00% 0 00% -0.05% -1 22%

Apr-17 (821 0731 0 (791) )1.847) -0 14% -0 77% 0 00% 0 00% -7 34% -0 46%

May-17 (1,(95) 39 (0) 478 (677) -I 62% 0 03% 0 00% 0 00% 4 31% -0 16%

Jun-17 1,784 1,288 0 0 407 3,479 1 69% 0 87% 0 00% 0 00% 3 86% 0.70%

Test Year Total (19,197) 97 0 O 692 (18,409) -1.79% 0.01% 0.00% 0.00% 0.55% -0.3%

Updated Test Year Total (16,561) 787 1 O 455 05,318) -1.53% 0.05% 0.00% 0.00% 0.36% -0.3%

EI9

JO

Z0

.17 - 1



New Mexico Weather Im acted Sales Calculation New Mexico Retail Weather Adjustment Summary

Heating Degree Days 10-yr nonnal Heating Degree Days Actual Vanance from Normal

Apr-16 124 137 13

May-16 31 62 31

Jun-I6 0 0 0

Jul-16 0 0 0

Aug-16 0 0 0

Sep-16 7 2 (5)

Oct-I6 129 35 (94)

Nov-16 448 349 (99)

Dec-16 749 732 (17)

Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-I6 Cooling Degree Days 10-yr normal 58 202 483 506 501 253 53 0 0 Cooling Degree Days Actual 20 158 478 672 427 231 78 2 0 Variance from Nonnal (38) (44) (5) 166 (74) (22) 25 2 (0)

Apr-16 May-16 Jun-16 Jul-16 Aug-I6 Sep-16 Oct-16 Nov-I6 Dec-16 Precipitation 10-yr normal 0 40 1 48 1 43 1 96 1 68 2 03 1 10 0 22 0.49 Precipitation Actual 0 77 0.60 0 53 0.45 5 05 1 21 0.46 0.71 0 30 Vanance from Normal 0 37 (0 88) (0 90) (I 51) 3 37 (0 82) (0 64) 0 49 (0 19)

Apr-I6 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-I6 Nov-16 Dec-16 New Mexico Residential Service (1) (5) (1,165) (227) 9,129 (3,974) (1,006) 418 (1,199) (370) New Mexico Residential Space Heat (2) 98 (689) (140) 5,961 (2,541) (576) (386) (2,281) (615) New Mexico Small General Service (3) (I) (173) (34) 1,352 (584) (149) 62 (117) (40) New Mexico Secondary General Service (4) & (5) (3) (701) (133) 5,121 (2,278) (632) 291 0 0 New Mexico lmgation Service (6) (350) 570 489 735 (3,062) 482 0 0 0 Large Municipal and School Service (7) (252) 179 0 0 0 0 0 0 0 Small Municipal and School Service (8) (28) 179 252 833 (464) (379) 298 0 0 New Mexico Retail Weather Adjustment Total (542) (1,799) 206 23,131 (12,902) (2,261) 684 (3,597) (1,025)

EI9

.10 £

017 -

1 W

I



New Mexico Weather lm acted Sales Calculation New Mexico Retail Weather Adjustment Summary


Jan-17 749 647 (102)

Feb-17 537 380 (157)

Mar-17 324 175 (149)

Apr-17 124 112 (12)

May-17 31 10 (21)

Jun-17 0 0 0

Test Year Total 3,097 2,519 (578)

Updated Test Year

Total 3,097 2,442 (655)

Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Total Total Cooling Degree Days 10-yr normal 0 0 5 58 202 483 2,061 2,061 Cooling Degree Days Actual 0 0 , 34 74 173 505 2,100 2,196 Vanance from Normal 0 (0) 29 16 (29) 22 39 135

Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Total Total Precipitation 10-yr nonnal 0 28 0 32 0.31 0 40 1 48 1 43 11.71 11 71 Precipitation Actual 1 09 0 16 0 32 1 45 0 54 0 56 11 65 12.30 Variance from Nonnal 0 81 (0 16) 0 01 1 05 (0 94) (0.87) (0 06) 0 59

Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Total Total New Mexico Residential Service (1) (2,150) (2,560) (1,070) 0 (752) 1,106 (4,179) (2,427) New Mexico Residential Space Heat (2) (4,151) (5,866) (3,832) (82) (443) 678 (15,018) (14,134) New Mexico Small General Service (3) (287) (453) (211) 0 (113) 169 (636) (371) New Mexico Secondary General Service (4) & (5) 0 0 0 0 (452) 645 1,666 2,695 New Mexico Irrigation Service (6) (24) 239 (11) (973) 605 474 (933) (1,537) Large Municipal and School Service (7) 0 0 0 0 0 0 (73) 0 Sinall Municipal and School Service (8) (9) 87 (5) (791) 478 407 765 455 New Mexico Retail Weather Adjustment Total (6,620) (8,552) (5,130) (1,847) (677) 3,479 (18,409) (15,318)

New Mexico Residential Service

Heating Degree Days Weather Adjustment (10-yr normal) Apr-16 May-I6 Jun-I6 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16

Vanance from Nonnal 13 31 0 0 0 (5) (94) (99) (17)

Heatmg Degree Days Weather Coefficients 0 0000000 0.0000000 0 0000000 0.0000000 0 0000000 0 0000000 0 0000026 0 0002074 0 0003754

Res Service Customers 58,781 58,824 58,801 58,757 58,777 58,690 58,602 58,616 58,640

Heating Degree Days Weather Adjustment (MWh) 0 0 0 0 0 0 (14) (1,199) (370)

Cooling Degree Days Weather Adjustment (10-yr normal) Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-I6 Nov-16 Dec-I6

(38) (44) (5) 166 (74) (22) 25 2 (0) 0.0000024 0 0004480 0 0008407 0.0009354 0 0009185 0 0007651 0.0002940 0.0000000 0 0000000

58,781 58,824 58,801 58,757 58,777 58,690 58,602 58,616 58,640

(5) (1,165) (227) 9,129 (3,974) (1,006) 432 0 0

(5) (1,165) (227) 9,129 (3,974) (1,006) 418 (1,199) (370)

Vanance from Normal Cooling Degree Days Weather Coefficients Res Service Customers Cooling Degree Days Weather Adjustment (MIMI)

NM Residential Service Weather Adjustment (1)

EI9

J0 1

7017

- 1 GIII



New Mexico Weather Impacted Sales Calculation

New Mexico Residential Service

Heating Degree Days Weather Adjustment (10-yr normal) Jan-17 Feb-17 Mar-17

Variance from Normal (102) (157) (149) Heating Degree Days Weather Coefficients 0.0003608 0.0002770 0 0001225 0 Res Service Customers 58,716 58,748 58,835

Apr-17 May-17 Jun-17 (12) (21) 0

0000000 0.0000000 0.0000000 58,795 58,862 58,880

Heating Degree Days Weather Adjustment (MWh) (2,150) (2,560) (1,070) 0 0 0


0 (0) 29 16 (29) 22 0 0000000 0.0000000 0 0000000 0 0000000 0 0004373 0.0008389

58,716 58,748 58,835 58,795 58,862 58,880 0 0 0 0 (752) 1,106

(2,150) (2,560) (1,070) 0 (752) 1,106

Vanance from Normal Cooling Degree Days Weather Coefficients Res Service Customers Cooling Degree Days Weather Adjustment (MWh)

NM Residential Service Weather Adjustment (1)

£I9

JO

gO

t - 1




New Mexico Residential Space Heat

Heating Degree Days Weather Adjustment (10-yr normal) Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16

0 0 (5) (94) (99) (17) 0 0000000 0.0000000 0 0000006 0 0002188 0 0007806 0 0012345

29,726 29,743 29,697 29,679 29,640 29,649 Heating Degree Days Weather Adjustment (MWh) 101 0 0 0 0 (0) (613) (2,281) (615)


Vanance from Normal Heating Degree Days Weather Coefficients Res Space Heat Customers

Apr-16 May-16 Jun-16 13 31 0

0.0002545 0 0000000 0.0000000 29,826 29,775 29,724

Apr-I6 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-I6 Nov-16 Dec-16 (38) (44) (5) 166 (74) (22) 25 2 (0)

0 0000029 0.0005237 0 0010234 0 0012074 0.0011607 0 0008660 0.0003046 0 0000000 0 0000000 29,826 29,775 29,724 29,726 29,743 29,697 29,679 29,640 29,649

(3) (689) (140) 5,961 (2,541) (576) 227 0 0

98 (689) (140) 5,961 (2,541) (576) (386) (2,281) (615)

Variance from Nonnal Cooling Degree Days Weather Coefficients Res Space Heat Customers Cooling Degree Days Weather Adjustment (MWh)

NM Residential Space Heat Weather Adjustment (2)

EI9

J0 90t - I C

M




New Mexico Residential Space Heat


Variance from Normal (102) (157) (149) (12) (21) 0 Heating Degree Days Weather Coefficients 0 0013781 0.0012564 0 0008688 0 0002374 0 0000000 0 0000000 Res Space Heat Customers 29,672 29,683 29,703 29,661 29,663 29,654 Heating Degree Days Weather Adjustment (MWh) (4,151) (5,866) (3,832) (82) 0 0


0 (0) 29 16 (29) 22 0 0000000 0 0000000 0 0000000 0 0000000 0 0005113 0 0010203

29,672 29,683 29,703 29,661 29,663 29,654 0 0 0 0 (443) 678

(4,151) (5,866) (3,832) (82) (443) 678

Variance froin Nonnal Cooling Degree Days Weather Coefficients Res Space Heat Customers Cooling Degree Days Weather Adjustment (MWh)

NM Residential Space Heat Weather Adjustment (2)

ET9

Jo

L01

7 -

I U

N




New Mexico Small General Service

Heating Degree Days Weather Adjustment (10-yr normal) Apr-16 May-16 Jun-I6 Jul-16 Aug-16 Sep-16 Oct-16 Nov-I6 Dec-16

Vanance froin Nonnal 13 31 0 0 0 (5) (94) (99) (17) Heating Degree Days Weather Coefficients 0 0000000 0 0000000 0 0000000 0 0000000 0 0000000 0.0000000 0.0000013 0.0001036 0 0002097 Small General Service Customers 11,362 11,381 11,438 11,464 11,495 11,511 11,502 11,479 11,481 Heating Degree Days Weather Adjustment (MWh) 0 0 0 0 0 0 -I -117 -40

Cooling Degree Days Weather Adjustment (10-yr normal) Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-I6

(38) (44) (5) 166 (74) (22) 25 2 (0) 0 0000019 0 0003443 0 0006551 0 0007101 0 0006906 0 0005796 0 0002203 0 0000000 0 0000000

11,362 11,381 11,438 11,464 11,495 11,511 11,502 11,479 11,481 (1) (173) (34) 1,352 (584) (149) 64 0 0

(1) (173) (34) 1,352 (584) (149) 62 (117) (40)

Variance from Normal Cooling Degree Days Weather Coefficients Small General Service Customers Cooling Degree Days Weather Adjustment (MWh)

NM Small General Service Weather Adjustment (3)

EI9

J0 8

0t - 1

MI




!New Mexico Small General Service


Variance from Norinal (102) (157) (149) (12) (21) 0 Heating Degree Days Weather Coefficients 0 0002457 0 0002501 0 0001234 0 0000000 0 0000000 0 0000000 Small General Service Custorners 11,493 11,505 11,509 11,509 11,553 11,567 Heating Degree Days Weather Adjustment (MW1i) -287 -453 -211 0 0 0


Vanance from Normal 0 (0) 29 16 (29) 22 Cooling Degree Days Weather Coefficients 0.0000000 0 0000000 0 0000000 0 0000000 0 0003361 0 0006535 Small General Service Customers 11,493 11,505 11,509 11,509 11,553 11,567 Cooling Degree Days Weather Adjustment (MW11) 0 0 0 0 (113) 169

NM Small General Service Weather Adjustment (3) (287) (453) (211) 0 (113) 169

EI9

J0 6

017

- I




New Mexico Small Secondary General Service


Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16

Variance from Normal (38) (44) (5) 166 (74) (22) 25 2 (0)

Cooling Degree Days Weather Coefficients 0.0000255 0.0046466 0.0084063 0.0089757 0.0090614 0 0082468 0 0033825 0.0000000 0 0000000

Small Secondary General Customers 3,397 3,413 3,429 3,434 3,414 3,420 3,428 3,433 3,448

Cooling Degree Days Weather Adjustment (MWh)

NM Small Secondary General Service Weather Adjustment (4)

-3 -701 -133 5,120 -2,277 -632 291 0 0

(3) (701) (133) 5,120 (2,277) (632) 291 0 0




U9,1

001

17- T

UN

New Mexico Small Secondary General Service


Vanance from Normal 0 (0) 29 16 (29) 22 Cooling Degree Days Weather Coefficients 0 0000000 0.0000000 0 0000000 0 0000000 0 0044567 0.0082643

General Custoiners 3,454 3,463 3.468 3,475 3,474 3,483

NM Small Secondary General Service Weather Adjustment (4)

Small Secondary Cooling Degree Days Weather Adjustment (MWh) 0 0 0 0 -452 645

0 0 0 0 (452) 645




EI9

J0 T

it -

I

Documents

RD 1 - 312 of 613