Balanced Supply And Demand: Analysis Of Current Working Gas Storage Data

Aug.31.14 | About: The United (UNG)

Summary

I'm comparing the current refill of working gas in storage so far from past years by imputed weeks.

The result shows that refilling working natural gas in storage is on pace to meet EIA's adjusted expectations.

2014 accumulative CDD is about 11.9% less than 5-year averages, which has been very helpful in rebuilding working gas storage.

According to US Energy Information Administration EIA's newly released report (Thursday, Aug 28, 2014), working gas (dry natural gas) in storage was 2,630 Bcf (Billion cubic feet) as of Friday, August 22nd, 2014. This represents a net increase of 75 Bcf from the previous week. Since the beginning of current injection season (April to October) the total net change of working gas in storage was 1,808 Bcf as shown in Table 1.

Table 1: Net Increase of Working Gas Storage Since Beginning of Injection Season

Week Ending

Storage -bcf-

Increase from Last Week

Mar 28, 2014

822

Apr 04, 2014

826

4

Apr 11, 2014

850

24

Apr 18, 2014

899

49

Apr 25, 2014

981

82

May 02, 2014

1,055

74

May 09, 2014

1,160

105

May 16, 2014

1,266

106

May 23, 2014

1,380

114

May 30, 2014

1,499

119

Jun 06, 2014

1,606

107

Jun 13, 2014

1,719

113

Jun 20, 2014

1,829

110

Jun 27, 2014

1,929

100

Jul 04, 2014

2,022

93

Jul 11, 2014

2,129

107

Jul 18, 2014

2,219

90

Jul 25, 2014

2,307

88

Aug 01, 2014

2,389

82

Aug 08, 2014

2,467

78

Aug 15, 2014

2,555

88

Aug 22, 2014

2,630

75

Total Increase:

1,808

Note: Compiled from EIA data (same data source for other tables and figure).

Week ending April 4 week was the first week of 2014 Injection Season.

Click to enlarge

In SHORT-TERM ENERGY OUTLOOK, released on August 12, 2014, EIA projects working gas storage of 3,463 Bcf at the end of October. Thus, in the remaining 10 weeks, we have a task of 833 Bcf to meet the goal.

Is the net change of 75 Bcf in the new release good or bad? Can we bring back the working gas stock to EIA's projected 3,463 Bcf or higher by the end of the current injection season? We can compare it to history for a suggestion even under a background of recently shale gas boom. In my Seeking Alpha article - Analysis Of Weekly EIA Working Gas Storage Data On Daily Basis - A New Look On Working Gas Storage Numbers, I transformed EIA's past weekly working gas storage data into a workable structure for 2014. The correctness of the transformed data on daily basis was validated by being able to derive the same 5-year moving average numbers as in EIA's 5-Year average spreadsheet. In this way we can compare the working gas storage numbers in 2014 on the same month and same day among different years more accurately.

I am going to compare the current working gas in storage numbers to the past. First, I picked up the recent years from 2007 to 2013 as shown in Table 2. The table shows that since the beginning of this injection season (April-04), the net change of storage beats the seven-year average in all weeks except the April 4, April 11 and July 25 weeks. So far, we have an accumulated net increase that's 384 Bcf more than the seven-year average. This is not bad. But keep in mind, the net injections in the last seven years were far from EIA's projected 2,637 Bcf as shown in Table 3. The average net injection of the last seven years was only 2,046 Bcf - that is 591 Bcf less than 2,637 Bcf projected by EIA.

Table 2. Comparison of Working Gas Storages by Week*

____2007 - 2013____

____2014______

7 Days Ending on

Average of Storage

Change from Last Week

Storage

Change From Last Week

Compare to 2007 - 2013

28-Mar

1,695

.

822

4-Apr

1,704

9

826

4

-5

11-Apr

1,731

27

850

24

-3

18-Apr

1,768

37

899

49

12

25-Apr

1,832

64

981

82

18

2-May

1,905

73

1,055

74

1

9-May

1,991

86

1,160

105

19

16-May

2,081

90

1,266

106

16

23-May

2,175

94

1,380

114

20

30-May

2,273

98

1,499

119

21

6-Jun

2,361

88

1,606

107

19

13-Jun

2,444

83

1,719

113

30

20-Jun

2,529

85

1,829

110

25

27-Jun

2,602

73

1,929

100

27

4-Jul

2,680

78

2,022

93

15

11-Jul

2,752

72

2,129

107

35

18-Jul

2,807

55

2,219

90

35

25-Jul

2,859

107

2,307

88

-19

1-Aug

2,910

51

2,389

82

31

8-Aug

2,953

43

2,467

78

35

15-Aug

3,003

50

2,555

88

38

22-Aug

3,064

61

2,630

75

14

Total

1,808

384

* EIA's weekly reports are ending Friday. This table uses 2014's Friday's months and days to select weeks in other years.

Click to enlarge

Table 3. Net Changes of Working Gas Storage in Injection Seasons from 2007 to 2013

Year

Net Change

2007

1,964

2008

2,172

2009

2,179

2010

2,217

2011

2,223

2012

1,486

2013

2,081

Average of 7 Years

2,046

Note: Compiled from EIA data.

Injection Season: April - October.

Click to enlarge

We could argue that since the beginning of the injection season, we have 384 Bcf more injection than we did through this point in the last seven years (Table 2) - if we could keep this rate of injection, we would fill a gap of 591 Bcf (2,637 - 2,046). I would like to crack the numbers further. I supplied Figure 1, which illustrates the net changes of working gas in storage by withdrawal and injection seasons.

Click to enlarge

From Figure 1 we notice that 2003 saw a net injection of 2,406 Bcf. That was close to EIA's projection number of 2,637 Bcf. Thus I picked 2003 to do a case study, which is summarized in Table 4.

Table 4. Comparison of Working Gas Storages by Week*

__2003__

___2014___

2014 Versus 2003

7 Days Ending on

Storage

Change from Last Week

Storage

Change from Last Week

Storage Change

Difference of Changes#

28-Mar

696

.

822

4-Apr

688

-8

826

4

138

12

11-Apr

642

-46

850

24

208

70

18-Apr

702

60

899

49

197

(11)

25-Apr

754

52

981

82

227

30

2-May

836

82

1,055

74

219

(8)

9-May

900

64

1,160

105

260

41

16-May

990

90

1,266

106

276

16

23-May

1,085

95

1,380

114

295

19

30-May

1,199

114

1,499

119

300

5

6-Jun

1,324

125

1,606

107

282

(18)

13-Jun

1,438

114

1,719

113

281

(1)

20-Jun

1,565

127

1,829

110

264

(17)

27-Jun

1,662

97

1,929

100

267

3

4-Jul

1,809

147

2,022

93

213

(54)

11-Jul

1,904

95

2,129

107

225

12

18-Jul

1,981

77

2,219

90

238

13

25-Jul

2,062

81

2,307

88

245

(70)

1-Aug

2,138

76

2,389

82

251

6

8-Aug

2,222

84

2,467

78

245

(6)

15-Aug

2,299

77

2,555

88

256

11

22-Aug

2,352

53

2,630

75

278

22

Total

1,656

1,808

152

* EIA's weekly reports are ending Friday. This table uses 2014's Friday's months and days to select weeks in 2003.

* 04-April as start week.

# Difference of the changes from the last week between two years.

Click to enlarge

Table 4 tells us that in the past 20 weeks, we've had 152 Bcf more refill than we did in 2003. In the rest of 10 refill weeks, we need a net increase of 833 Bcf to meet EIA's target, averaging about 83 Bcf/week. As September is a month typical with larger injection figures, I hereby confirm my previous opinion that we are on the way to normalizing the depleted working gas in storage if the current refill momentum is maintained or improves. This point of view is also strengthened by the current EIA's Natural Gas Weekly Update Supply and Demand tables:

U.S. Natural Gas Supply - Gas Week: (8/20/14 - 8/27/14)

Percent change for week compared with:

last year

last week

Gross Production

6.49%

0.68%

Dry Production

6.43%

0.68%

Canadian Imports

-1.97%

-0.95%

West (Net)

-24.26%

-4.33%

MidWest (Net)

46.86%

1.40%

Northeast (Net)

-990.77%

-7.06%

LNG Imports

-70.62%

6.09%

Total Supply

5.45%

0.57%

Source: BENTEK Energy LLC

U.S. Consumption - Gas Week: (8/20/14 - 8/27/14)

Percent change for week compared with:

last year

last week

U.S. Consumption

2.7%

2.6%

Power

1.7%

7.5%

Industrial

1.1%

-0.9%

Residential/Commercial

8.9%

-3.8%

Total Demand

2.2%

2.0%

Source: BENTEK Energy LLC

Click to enlarge

Some Seeking Alpha readers were against my model of comparison to 2003's data as 2003 was a pre-shale year without robust NG output increase overtimes. But I think this point of view can only support my model. In the rest of 10 weeks towards the end of October in 2003, we had a net injection of 803 Bcf as shown in Table 5, which is very close to 833 Bcf that we need for 2014, to meet EIA's target. Please bear in mind, in 2003, there was a very high September's and an average October's compared to 2000-2013's injection statistics (Table 7).

Table 5: Net Increase of Working Gas Storage in the Next 10 Weeks of 2003 Injection Season

Week Ending on

Storage

Change from Last Week

29-Aug

2,419

67

5-Sep

2,518

99

12-Sep

2,619

101

19-Sep

2,719

100

26-Sep

2,820

101

3-Oct

2,904

84

10-Oct

2,981

77

17-Oct

3,066

85

24-Oct

3,121

55

31-Oct

3,155

34

Total:

803 -bcf-

Table 6. Comparison of Accumulated Cooling Degree Day (CDD)

Accumulated CDD since Beginning of Injection Season

7 Days Accumulated CDD

2014 Week Ending

5 Year Average

2014

% Change

5 Year Average

2014

% Change

4-Apr

4.2

5

19

6.2

6

-3.2

11-Apr

12.2

12

-1.6

8

7

-12.5

18-Apr

19.8

17

-14.1

7.6

5

-34.2

25-Apr

30.4

24

-21.1

10.6

7

-34.0

2-May

46

38

-17.4

15.6

14

-10.3

9-May

64.2

59

-8.1

18.2

21

15.4

16-May

84.6

87

2.8

20.4

28

37.3

23-May

110.6

108

-2.4

26

21

-19.2

30-May

152.4

146

-4.2

41.8

38

-9.1

6-Jun

197.6

191

-3.3

45.2

45

-0.4

13-Jun

247

234

-5.3

49.4

43

-13.0

20-Jun

303.2

292

-3.7

56.2

58

3.2

27-Jun

375.6

353

-6

72.4

61

-15.7

4-Jul

448.6

426

-5

73

73

0.0

11-Jul

531.2

496

-6.6

82.6

70

-15.3

18-Jul

615

560

-8.9

83.8

64

-23.6

25-Jul

701.4

628

-10.5

86.4

68

-21.3

1-Aug

778.9

691

-11.2

77.8

64

-17.7

8-Aug

859.2

755

-12.1

81.2

64

-21.2

15-Aug

933.6

815

-12.7

74.4

60

-19.4

22-Aug

1,003.60

884

-11.9

70

69

-1.4

Note: Accumulated CDD is calculated by adding daily CDD since April 1st or the last seven days.

65°F is used as base temperature.

This table is compiled from NOAA Degree Days Statistics database.

Click to enlarge

This week's number (75 Bcf) is little below my expectation, which may be due to the last week's normal CDD to 5-year average's (Table 6). Can we have a favorable weather and an about 5% dry natural gas output increase year/year in the rest of 10 injection weeks? This is a question only time will tell for sure, but I hold positive for the natural gas supply part. Please note the EIA's target of 3,431 Bcf working gas in storage is about 11% below the five-year average (3,842 Bcf). This would support the next-month working natural gas future price above $3.7/mmBtu in October. I advise readers to be cautious in trading United States Natural GAS (NYSEARCA:UNG) and other related securities, especially holding short positions around $4/mmBtu. But I don't see an immediate buying opportunity either because I'm not convinced that $4.5/mmBtu could be reached in short time. Experienced traders may make short-term, only long speculative bets on fluctuations around $4/mmBtu based on next-month natural gas price though. In my opinion, prices could be grinding a little higher than $4 and have little chance to drop below $3.7/mmBtu and stay there even for a few weeks by the end of October, unless supply increases unexpectedly (>= 6 bcf/day net increase over 2013). Actually we have had a cooler than average summer so far. From Table 6, we notice that there is about 11.9% less accumulative CDD (US continental, population weighted) in 2014 compared to its five-year (2009-2013) average. This undoubtedly has contributed a lot to 2014 working gas storage build-up, as natural gas fired turbines are not called upon as often to respond to heat waves.

Lastly, I prepared Table 7 that may be useful for readers interested in trading natural gas related securities. Readers with less statistical knowledge may only want to pay attention to mean, minimum and maximum though. Some SA readers kindly pointed out that Table 7 failed to include robust 2014 natural gas output increase. Author therefore agrees that about 20 Bcf/week more could be added to building your model when referencing Table 7.

Table 7. Simple Statistics of Natural Gas Injection Amount by Months in Injection Seasons from 2000-2013

Mean

Standard

Standard

95% Confidence Interval

Month

(Average)

Minimum

Maximum

Median

Deviation

Error

Low Limit

High Limit

April

192

46

360

197

80

21

146

238

May

387

231

479

402

69

18

347

427

June

354

232

469

342

62

16

318

389

July

281

134

385

282

76

20

237

325

August

244

127

350

240

69

18

204

284

September

328

269

411

313

47

12

301

355

October

239

87

385

245

79

21

194

285

Sum of 7 Months

2,025

1,126

2,839

1,747

2,303

Note: Injection Season: April - October.

95% confidence interval is computed assuming net changes (Injection amount) are normally distributed over years by months in inject season and should be interpreted with caution.

Click to enlarge

I am keeping a close watch on the progress of working gas in storage refill and will share my research with Seeking Alpha readers accordingly.

Disclosure: The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: These are just my opinions without any third party interest. Readers should be aware of high investment risk in energy sectors, especially in coal and natural gas industries, which are highly speculative. Readers should consult their own financial advisers before making any investment decision and trading. It is to the best of author's effort to collect and further calculate EIA's data in this article with accuracy. Readers should take due diligence in reading all numbers in this article and are not encouraged to use these numbers in any purpose without their own validation.