From ~104 bcf/d in February to an average in May so far of just ~99 bcf/d, L48 gas production has fallen ~5bcf/d in an astonishingly short period of time. 75% of that decline has come from the NorthEast and Haynesville - LA where anecdotes of shut ins, deferred TILs and productive capacity decline dominate the narrative around trying to explain exactly how this happened. Categorizing the production decline into these three buckets helps us better understand how long production could remain subdued, the circumstances under which it could return and how quickly we can expect it to grow given those circumstances. There is still much that we dont know, but between the available state data, SynMax proprietary data, including our TIL investigation, and what the producers themselves are saying we can begin to get our arms around these very important questions. We will do our best throughout this report to clearly differentiate fact, extrapolation and speculation.
Three Buckets
Productive Capacity Decline
Speed of Return - Frac Crew Addition + 60 days
This category represents the amount of production lost due to decreases in fracking/completion activity. Loss of production from this category is limited by the vintage declines of the region. Regaining lost production in this category is limited by the ability to increase frac crew counts. Historically we have observed peak frac crew additions across all regions at a rate of one crew per day. Once in operation, a crew's influence on producing capacity is lagged by 60 days.
Deferred TILs
Speed of Return - 30 days
This category represents production missing due to deliberately deferred TILs. NYT (not yet TILd) inventories build slowly as otherwise regularly scheduled TILs add to a portfolio of new production which can be brought online within 30 days according to producer estimates.
Shut Ins
Speed of Return - 0 to 7 days
Sometimes referred to as curtailments, this category represents existing production deliberately taken off line. Production in this category can change quickly, sometimes within a day as producers in many regions can remotely turn wells off and on.
Haynesville - LA
Lost ~1.25bcf/d - Mostly Prod Decline, but some DTILs
The Haynesville - LA decline from 11.25 bcf/d to most recently 10 bcf/d follows a lagged decline in fracking activity which began in December reducing crews by more than 50% to its low point in February. The short term forecast, an unconstrained model which at present does not incorporate the impact of DTILs or curtailments, models 0.90 bcf/d of this roughly 1.25 bcf/d drop.
As noted in our TIL investigation reports, NYT inventories seem to be 85% a NorthEast phenomenon, with only marginal DTILs in the Haynesville, according to our sample. This makes sense given the NorthEastern gas price discount to the hub.
Haynesville production is declining as expected given the decline in fracking activity. This is shown by production’s close adherence to our Short Term Forecast. There is however a gap of 0.3 - 0.4 bcf/d which we believe is explained by DTILs.
SW PA
Lost ~1.25bcf/d - Mostly Shut Ins
From a mid-February rate of 9.85 bcf/d to the recent rate of 8.6 bcf/d most of the SW PA decline came seemingly at once with a sudden drop of 1 bcf/d over just two days, Feb 23-25. This by itself is characteristic of producer shut ins but we can further confirm this story with well level state data which is now 85% complete for March.
Shut-in wells according to state data. Operator field is displayed as reported, so RICE is now EQT.
The combination of shut-in like pipe scrape behavior and an apparent 1.1 bcf/d of sample adjusted state data showing the same between Feb and March leads us to believe that in SW PA the bulk of the prod decline is coming from producer curtailments.
NE PA
Lost ~1.25bcf/d - Mostly DTILs
From an early February rate of 12.15bcf/d to the most recent readings of around 10.9 the decline profile, while lumpy, is far more gradual than the SW PA sudden drop indicative of well shut-ins. As described above, our analysis of well level data, which is now 85% complete for the month of March, shows only a small portion of the 0.934 bcf/d of shut ins confirmed between Feb and March belongs to NE PA. Just 0.09 bcf/d or 0.105 bcf/d when grossed up. This also fits what we have observed in our TIL investigation, where our study of just one producer, CHK’s NYT inventory revealed 0.76 bcf/d of DTILs in NE PA.
CHKs NYT Inventory in NE PA.
The low confirmed shut-ins appearing in the Feb to March state data along with the high concentration of CHK DTILs and the more gradual decline profile of NE PA leads us to believe that this region's decline is primarily a DTIL story.
The Implications
We believe the currently available evidence supports the idea that production declines are coming in equal parts from DTILs, Shut Ins and Productive Capacity Decline. The implications are that should prices rise to the satisfaction of producers only 1.25 bcf/d of shut-ins can be brought immediately back, 1.25 bcf/d of DTILs would come next on a 30 day lag. The return of productive capacity would take several months or more as crews must first be mustered back into service before they could have a 60 day lagged impact. The most important implication of this theory is that the bulk of quickly available production is stuck in the NorthEast where the benefit of this most recent hub rally has been muted. This could delay a meaningful return of production to November if producers decide to take maximum advantage of NorthEast contango.
Our view is subject to change, as always, should evidence emerge that our estimation of shut-ins, DTILs and productive capacity is incorrect. Our DTIL coverage in the Haynesville and NorthEast is set to further expand in the coming weeks, new state data is constantly becoming available and producer earnings offer glimpses into activity that could challenge our theory. We intend to stay on top of this developing narrative and incorporate new information in real time for the benefit of our customers. As always please reach out to us with any questions.