Next week's Panorama will be released on Wednesday 21 November, due to the Thanksgiving holiday
Today’s report (week ended 9 Nov): EIA: +39 bcf, EA: +39 bcf
- Today’s EIA print — while in line with our forecast — showed a wide divergence between estimates, especially between S&D and flow models, which had a much lower bias.
Next Wednesday’s report (week ending 16 Nov): EA preliminary: -122 bcf
- A nearly two-fold w/w increase in GWHDDs is driving next week’s withdrawal, with res-com ramping up by 15 bcf/d w/w and power by 5 bcf/d w/w. Maintenance activity, including in the Permian, led to a 0.5 bcf/d w/w production decline that was more than offset by an uptick in Canadian trade.
A freeze-off threshold
Cold, both current and future, has been driving the market. National GWHDDs for the week in progress are comparable to those in place for the week ending 22 December 2017. Unseasonably low temperatures are spurring the first triple-digit withdrawal of the heating season for next Wednesday’s EIA report. However, this early cold has not been chilly enough to spur any production losses. The losses that we have seen―concentrated in the Permian, where evening lows have hit below freezing in the past two days―have stemmed, in part, from a rupture on the Sand Hills NGL line and maintenance on Northern Natural Gas (see Fig 3). With some commercial weather forecasts calling for a cold December as well, the spectre of a proper bout of production freeze-offs is looming. With this in mind, this week’s Panorama examines the weather conditions that spurred output interruptions last year.
Right now, based on both our pipeline flow sample and forecast temperatures in key basins, freeze-offs are not currently likely to remove any volumes from production. Daily output in the Permian has historically only been significantly disrupted when temperatures have fallen below 20°F for two consecutive days (see Fig 1). West Texas dipped below freezing on the evenings of 13 and 14 November, but the forecast through Thanksgiving is for lows near 40°F. The cluster of Permian points in Fig 1 show even average low temperatures of 25°F or above tend not to affect regional output. Appalachia has a similar threshold for freeze-offs—sustained low temperatures of around 20°F. However, in the event of temperatures below the threshold, there could be higher outright lost volumes from both Appalachia and Permian than previous weather events given that both regions combined have a higher output baseline on the order of 5 bcf/d y/y. The potential exists for d/d losses of more than 1.5 bcf/d in Appalachia alone if temperatures below that 20°F threshold hit the Northeast, as Fig 1 shows was the case during the January 2018 cold snap. The term freeze-off is most appropriate for the Haynesville, as consecutive days below 32°F typically cause lost volumes of 0.5-1.0 bcf/d. While not included in the chart (to keep the scale informative for the other regions), the low temperatures that typically coincide with lost production in the Bakken are closer to -20°F. Fundamentally, our view on balances has not changed from yesterday, and yet trading suggests a much different view. Near-term changes since yesterday in the 15-day forecast have been minimal, though GWHDDs for the week ending 30 November came off by 3%. Still, our balances are guiding for a 3.0 tcf end-November carryout.
As much as the run up in prices in the last two trading sessions had caught the market’s attention, so too has today’s drop (and subsequent limit down) in trading. As we mentioned yesterday in our E-mail alert: Push it to the limit (trading): Henry Hub flirts with five-handle pricing, 14 November 2018, some price action has been driven by positioning, short-covering, stop-outs and the follow-through of algorithmic trading on the movements of both of the former. Certainly, today’s ‘bearish’ miss versus consensus was worth a few pennies, but not a few dimes, especially with no notable change in fundamentals since yesterday. Trading volume today now stands at a fraction of what it was in the past several sessions as well (see Fig 4). Our forecasts do not peg another triple-digit withdrawal until the week ending 14 December. The next several reports will show a notable step down in storage activity, on lower GWHDDs and Thanksgiving holiday impacts. Subsequently, some step down in cash pricing has been anticipated on the lower call on storage (see Fig 5) that we identified in yesterday’s alert as not meaningfully below the high $3.00s/mmbtu, unless a significant return to milder weather emerges in the forecasts.
|Fig 1: Three-day average low temperature (°F) vs production loss from pre-freeze baseline, bcf/d|
|Note: Observances are from December 2017 and January 2018.
Source: Accuweather, Energy Aspects
|Fig 2: Appalachia output, bcf/d||Fig 3: Permian output, bcf/d|
|Source: Ventyx, Energy Aspects||Source: Ventyx, Energy Aspects|
|Fig 4: CME trading volume by month||Fig 5: Weekly storage activity, bcf|
|Source: CME, Energy Aspects||Source: EIA, Energy Aspects|