American Oil and Gas Reporter - May 2017 - 47

SpecialReport: Market In Motion

End-Of-Season Inventory
Although 2016-17 winter weather was again mild, strong exports and lower production allowed the market to withdraw nearly 2 Tcf between Nov. 1 and March 31. With 2 Tcf in storage at
the start of the 2017 injection season this spring, how likely is
it that the storage inventory will again approach or exceed the 4.0
Tcf level by the start of the next winter heating season on Nov.
1, as it did the past two years? Based on our scenarios, not very.
The dashed lines on the right side of Figure 4 project inventory through October based on the modeled supply/demand and
injection scenarios. Two of the projections assume U.S. consumption will drop to the three-year average (2.4 Bcf/d lower than 2016).
One assumes production will rise modestly to 2016 levels, but
exports will remain flat (orange dashed line). The other surmises that production will rise, but exports also will grow (blue dashed
line). The resulting season-ending inventory projections range from
3.5 Tcf with incremental export demand to 3.8 Tcf without it. In
all cases, the storage inventory at the end of the 2017 injection
season is projected to be well below 4.0 Tcf.
If consumption during this year's injection season tracks 2016's
record level, the implied injections will leave inventory close to

FIGURE 4
Gas Storage Inventories
(Historical versus 2017 Modeled Scenarios)
4,500

4,200
3,900
3,600
3,300
Bcf

3,000
2,700
2,400
2,100
1,800
1,500
Ju
l-1
5
Se
p15
No
v-1
5
Ja
n16
Ma
r-1
6
Ma
y-1
6
Ju
l-1
6
Se
p16
No
v-1
6
Ja
n17
Ma
r-1
7
Ma
y-1
7
Ju
l-1
7
Se
p17
No
v-1
7

1,200
Ja
n15
Ma
r-1
5
Ma
y-1
5

above the five-year average.
Next, if we look at scenarios using three-year average U.S. consumption volumes (orange bars), injections will range from 1.5
Tcf to 1.8 Tcf at most. That is well below the five-year line, but
potentially above 2016. Injections will land on the higher end of
that range if no incremental demand materializes (left-most orange bar), and even more so if production rises to 2016 levels (middle orange bar). But they will come in on the lower end of the
range (nearly flat with 2016) with 0.6 Bcf/d in export growth, even
if production also grows (right-most orange bar).
If 2017 injection season consumption averages the 2016 level (green bars), which is essentially also the three-year high for
seasonal demand, total injections will drop to near the 1.0 Tcf
mark, well under both the five-year and year-ago levels. More
specifically, it will be closer to 1.1 Tcf with no incremental export demand (left green bar) and slightly lower (1.0 Tcf) with export demand growth. And that will be true even if production returns to the 2016 level (right green bar).
Altogether, these scenarios suggest that unless production climbs
substantially or U.S. consumption falls drastically from last year,
injections are likely to be on the lower end of the spectrum. The
bottom line for price direction, however, is where these injection
scenarios leave the overall U.S. storage inventory next fall. To provide answers, inventory levels are projected through October based
on the different modeled injection scenarios (Figure 4).
As the blue solid line shows, storage inventory started the 2015
injection season at the relatively low level of 1.5 Tcf (blue oval),
but as production climbed, the market injected nearly 2.5 Tcf that
summer. By November, storage had peaked above 4 Tcf for the
first time (yellow oval). That record level was followed by the
mild 2015-16 winter, which left the year-over-year inventory much
higher at the start of the 2016 injection season (green oval).
However, record demand through much of the 2016 injection
season suppressed injections to less than 1.5 Tcf. While the inventory peaked at a new record of 4.047 Tcf in November 2016
(purple oval), it was not by much considering how high the inventory had been seven months earlier.

Total Lower 48

LY demand

3-yr avg dem + 2016 prod + exports growth

3-yr avg dem + 2016 prod

3-yr low demand

LY dem + 2016 prod + exports growth

3.0 Tcf by autumn (the green and black dashed lines). In fact, the
only modeled scenario that even approaches 4.0 Tcf is the lowdemand case, in which U.S. consumption declines to a three-year
low and exports do not grow (red dashed line). In that case, the
inventory will end well above 4.0 Tcf. This is the most bearish
scenario, but also the least likely, given the substantial new export capacity available heading into this injection season. In addition, if inventory begins to balloon, prices likely will be pressured to the point at which gas displaces more coal in the power sector, thereby stimulating consumption.
Considering that most of the scenarios lead to relatively bullish outcomes compared with last year, it is no wonder that $3.00
an MMBtu pricing has held in the futures curve, even as weather remains mild. That said, however, some curve balls still could
knock the market off its generally bullish course. For instance,
on the bearish side, production could roar back faster than expected in the second half of 2017. On the bullish side, exports or U.S.
consumption also could grow more than assumed. What we do
know from modeling these scenarios is that the year-on-year deficit
in storage and the underlying supply/demand balance give the bulls
an edge as the injection season gets into full swing.
❒

SHEETAL
NASTA

Sheetal Nasta is a fundamental energy analyst, writer
and consultant with a decade of experience observing and
analyzing natural gas markets. She is a regular contributor
to RBN Energy LLC's "Daily Energy Post" articles. Nasta
previously served as manager of energy analysis for the
North American natural gas market at Bentek Energy, and
as a senior editor on the natural gas desk at Platts. She holds
a bachelor's from Southwestern University and a master's
from Northwestern University.
MAY 2017 47

American Oil and Gas Reporter - May 2017

Table of Contents for the Digital Edition of American Oil and Gas Reporter - May 2017