« Back to Get the Point main page

If You Don’t Eat Your Meat, How Can You Have Any Pudding?

Power Burn’s Precarious Position in the Summer 2016 Gas Market

July 21, 2016 | By Jack Weixel

To date this summer, natural gas delivered to the electric power sector (power burn) has averaged higher than any previous summer on record. And not by a little – from April 1 to July 19, power burn is averaging 28.6 billion cubic feet per day (Bcf/d) and makes up 47% of domestic natural gas demand. This compares to last summer to date, when power burn averaged 26.2 Bcf/d and accounted for 44% of domestic natural gas demand. Compared to the five-year average, to date, power burn is up nearly 4.5 Bcf/d this summer, or 19% vs. the average in the first half of summers 2011 to 2015. 

But what does that mean for the gas market overall this summer?

In this week’s Get the Point, we take a look at the bigger picture, and it’s not quite as positive from a demand perspective as the record power burn implies. While the increase in demand in the power sector is impressive, in order to reach any semblance of price-supportive equilibrium, an even greater contribution on the demand side of the equation is needed. 

Production levels have not shown enough evidence of decline through mid-summer to reign in historically high storage inventories. For this reason, as well as others, PointLogic believes that the market is headed to another record storage inventory carryout level by the end of October. 

Below is a graph comparing month-on-month power over the past five summers.

Power Burn by Month (Bcf/d)

Source: PointLogic Energy Supply & Demand

Power burn has set a monthly average record this year in April, May and June. It’s averaged 1.5 Bcf/d more than those three summer months of 2012. June 2016 power burn beat June 2012 power burn by an astounding 3.3 Bcf/d. (Surprisingly, this trend has not continued into July to date, when power burn limped out the gate at the start of the July 4 holiday weekend, and is only now just recovering.)  

Certainly, power burn levels so far this summer have afforded producers the ability to drill for more gas and stave off bankruptcy (with several notable exceptions). But looking at natural gas storage injection rates, the amount of natural gas consumed for power has not been enough to bring inventories back within the five-year range, or even below the five-year maximum inventory level.

Taking a look at last year’s injection rates for the summer week-by-week compared to this summer’s rate yields the following graph.

Storage Injection Comparison (Bcf)

Source: EIA and PointLogic Energy Supply & Demand

It’s notable that in summer 2015 to date, the lower 48 U.S. experienced four weeks when injections topped 100 Bcf. This summer, the highest injection week was a mere 82 Bcf experienced during the week ending May 26.

Prior to and since then, injections have not even sniffed at a triple-digit injection. Similar to the summer of 2012, this summer will likely go down as one devoid of 100+ Bcf injections. 

On a weekly basis, storage injections this summer have been 35.8 Bcf lower than last year's levels, an impressive 5.1 Bcf/d. On a daily basis, power burn differentials during each week have made up about 2.4 Bcf/d of the summer on summer shortage. An increase in Mexican exports, LNG exports and a drop in production make up the rest. Still, storage inventories remain at all-time highs, clocking in at 3,243 Bcf for the week ending July 7, or 507 Bcf higher than last year’s levels. 

Power burn’s contribution to lower storage injections and lower storage inventories this year is real and significant. To start the storage injection season, inventories were a staggering 1,008 Bcf above last year. The fact that over 500 Bcf of gas has been cut out of the year-on-year surplus in just 14 weeks this summer is impressive, but the question remains: Could power burn be doing more? 

Start a Free Trial of PointLogic Energy's Pipeline Flow Data
As British schoolchildren and Pink Floyd know alike,” How can you have any pudding, if you don’t eat your meat?” Surely, power burn is doing as much as weather and price will allow, but the slowdown in July power burn levels means that it has that much more meat to eat over the course of the next 17 weeks of summer injection season. Should that meat not be eaten, production will likely have to make up the difference and decline at a faster rate.    

PointLogic certainly expected power burn to do more, as discussed in a Get the Point article a few weeks ago. Our article noted that maximum demonstrated power burn levels for the month of July could mean a potential average burn of 38.6 Bcf/d. Compared to last year, that would take a massive 5.2 Bcf/d of gas out of the surplus, equivalent to storage injections for July declining by over 161 Bcf compared to July 2015, in one month alone. 

PointLogic’s June edition of the Supply and Demand Two Year Balanced Forecast (available to subscribers here) hedges this high power burn number a bit, estimating that power burn would average 35.7 Bcf/d for July. That would still account for an extra 2.3 Bcf/d of incremental demand compared to July 2015. Alas, that has not been the case thus far in July, with power burn averaging only 34.5 Bcf/d through July 19. 

By mid-July, however, the relatively weak July numbers started to turn around as temperatures reached higher marks across the country. On July 13, power burn hit 37.6 Bcf/d, the fifth-highest daily burn rate in recorded history, and then it was surpassed on July 18, at 37.7 Bcf/d. Since July 5, eight of 14 days have recorded burn levels 36.0 Bcf/d or higher.

However, the fact remains that in order for July power burn to average 35.7 Bcf/d for the whole month, the remaining 12 days in July (as of this writing) would need to average 37.6 Bcf/d (equivalent to the highest daily burn rate seen this month and fourth highest on record). This includes four more weekend days, when power burn has only averaged 32.1 Bcf/d over the past three weekends. In order for power burn in July to average 38.6 Bcf/d in the month, the remaining 12 days would have to average 44.5 Bcf/d, or 0.9 Bcf/d above the highest daily maximum demonstrated power burn rate during any one day in July ever. 

While a “heat dome” is expected to envelop the middle of the country over the next few days, PointLogic does not see those records being set. Power burn is forecast to average 35.4 Bcf/d for the 13 days of July 19 through July 31. Also, this heat dome distinctly affects the middle of the country, an area that has some significant gas-fired power generation in operation, but less population to support than regions like the East Coast or Southeast.

 Highest Realfeel

Source: Accuweather

While June power burn was remarkable, beating June 2012 power burn by an average of 3.3 Bcf/d, PointLogic sees July 2016 power burn only modestly topping 2012. July 2016 is on track for a record, but only by 0.8 Bcf/d compared to July 2012 and 2.0 Bcf/d compared to July 2015.  

July Burn per Degree (Bcf/d versus Temp)

Source: PointLogic Energy

As the graph above indicates, July 2016 power burn tracks very similarly to July 2015 on a burn-per-degree basis. But when compared to July 2012, at 80 degrees Fahrenheit, July 2016 burn is up about 2.0 Bcf/d. This indicates that there is more natural gas capacity to burn in July 2016 and July 2015, compared to July 2012 when higher temperatures occur. This reflects the buildout in gas-fired capacity over the last several years.

At temperatures below 77 degrees, the trend line indicates that July 2012 power burn is slightly higher than both July 2016 and July 2015. Something else is afoot – the price of natural gas becomes more important.

During the first week of July 2012, the Henry Hub cash price averaged $2.84 per MMBtu, while during the first week of July 2016, Henry Hub averaged $2.89. While those prices are basically the same, the price of coal (a shrinking but still competitive fuel source for power burn generation) has become more competitive. Central Appalachian coal prices have sunk by nearly $20 per short ton (33%) from July 2012 to July 2016.       

As average temperatures and burn rates recede through the month of August, there is significant evidence to suggest that if current spot prices remain above the average price of $2.84 per MMBtu seen in August 2012, then this year's power burn on a degree-day basis, might not be incentivized to alleviate the supply surplus and record storage inventory overhang. 

The current price level of Henry Hub cash and the forward curve for the balance of summer (at $2.71 per MMBtu), while painful for producers, has not been painful enough to cause a massive slowdown in production. Production this summer to date is off just 0.7 Bcf/d compared to summer 2015 to date.

Lower 48 Gas Production 2016 vs. 2015 to date (Bcf/d)

Source: PointLogic Energy Supply & Demand

To further exacerbate the supply situation in the lower 48, the oversupply situation in Canada is forcing producers there to discount their gas to move into the U.S. Net Canadian exports to the U.S. are up nearly 0.5 Bcf/d this summer to date vs. summer 2015 to date, effectively negating the reduction in domestic production.   

With the current pace of injections reflecting only a marginal reduction on the supply side, paired with a steady increase in demand from power, Mexican and LNG exports summer on summer, PointLogic is projecting that the market will still have a glut of gas to deal with at the end of summer injection season. Below is a look at the summer to-date demand comparison to last summer to-date.  

Summer 2016 vs. Summer 2015 td (Bcf/d)

Source: PointLogic Energy Supply & Demand

The inventory level at the end of summer at the current pace, and given the demand constraints observed thus far, would result in the market being short gas about 3.3 Bcf/d. Over 214 days of summer, this equates to a total summer injection number that is 706 Bcf less than last summer. Last summer, injections from April 1 through the week ending Nov. 4 totaled 2,498 Bcf. Subtracting 706 Bcf would mean summer 2016 injections total only 1,792 Bcf. The level of inventory on April 1, 2016 was 2,480 Bcf, which implies that on Nov. 4, 2016, the lower 48 could see inventories at 4,272 Bcf. This would be a new inventory record, surpassing last year’s record of 4,006 Bcf set during the week of Nov. 18 by 266 Bcf. 

As Pink Floyd mused on Another Brick in the Wall, should power demand not finish its meat, there will be no pudding in the form of lower storage inventories and price support going into winter 2016/17. The implied surplus of around 266 Bcf year-on-year at the beginning of November means that the market still has long way to go to finish all that is on its plate. 

Could the market absorb this 266 Bcf surplus and start the winter season at the same level as last year?

As the number of days in summer becomes less, the burden for power burn becomes greater. With only 122 days left between July 19 and Nov. 18, power burn would have to eat a significant pile of meat in addition to the serving size last year to erase the 266 Bcf projected surplus at current pace.  

Power Burn by Month

Source: PointLogic Energy Supply & Demand

Splitting the 266 Bcf surplus evenly over the course of the 122 days through Nov. 18 means that an incremental 2.2 Bcf/d per day needs to be burned above the same time period last year. The last 12 days of July last year averaged an impressive burn rate of 35.7 Bcf/d. The current PointLogic projection for the next 12 days of July is an average of about 37.0 Bcf/d, or 1.3 Bcf/d higher, but 0.9 Bcf/d short of the 2.2 Bcf/d incremental bogey needed. 

This leaves the lion’s share of the power burn gap to be made up in August, September and October. In the above graph, the red bars represent the projected amount of power burn necessary each month to erase the storage surplus vs. last year.  August power burn would need to eclipse July 2012 burn levels, while September power burn would need to eclipse the highest June power burn on record, an event that has never occurred, and October power burn at the height of the shoulder season would need to eclipse the May production average, another event that has never occurred.

In other words, it's an unlikely scenario. But it’s important to note that the graph reflects only the impact on the power sector on the demand side and assumes no changes to the current amount of daily production, Canadian imports, Mexican exports or LNG exports. 

While the increased use of gas for power burn has shifted the demand curve upwards, it's not sufficient to achieve the record numbers needed without help from weather or prices. In order to facilitate this projected power burn profile, weather must be extremely warm, and prices must decline. Otherwise, production must decline in order to avoid further price erosion and a storage inventory crunch of unprecedented proportions. 

Stay tuned for more analysis of supply and demand factors leading up to the end of summer injection season.

« Back to Get the Point main page


Sign up here to have Get the Point delivered to you each week!





Share With Your Colleagues






© PointLogic Energy, part of OPIS an IHS Company | Site Map