Permian Takeaway Constraints and US Crude Outlook

During the past year, we have endured the ubiquitous Permian crude production versus pipeline takeaway charts, showing forecast crude output running well above anticipated pipeline capacity.  These charts may highlight the obvious shortfall in takeaway capacity, but the physics is dubious.  Although pipeline operators can take actions to boost pipeline throughput above the stated capacity, and are certainly scrambling to add any capacity to their Permian systems, crude outflows from the region are still limited to something near these capacity figures.  Regional pricing differentials are now signalling full pipelines, with the negative spread between WTI Midland and WTI Houston sliding towards the US$21/bbl record lows seen in August 2014.  With meaningful pipeline capacity unavailable until mid-2019, Permian production will need to flatten over the next 12 months.

Estimates for rail and trucking takeaway capacity from the Permian Basin vary, but the consensus is that these transportation modes are limited to only 100-150 kbpd.  The following chart shows our Base Case assumptions for rail and truck movements, versus the required volumes, based upon production and pipeline capacity forecasts.  The need for these outlets is short-lived, and disappears once the new pipelines come on line.  This timeframe is too short for producers to commit to the longer take-or-pay contracts that railroads would demand for the same reason, so this approach would remain limited.

This places significant pressure on Permian producers to limit production without an outlet.  One approach is to stop adding oil rigs to the basin.  The pace has certainly slowed during the past few weeks, averaging a little more than one a week since June, but the rig count needs to decline soon.  The following chart provides our Base Case forecast for oil rigs in the five major LTO basins, and suggests that Permian rigs need to move towards 430 by mid-2019, from the current 486 count.  Oil service companies are suggesting that the Eagle Ford and other LTO basins may benefit from rig redirection from the Permian.  Rig counts rebound in the Permian once the new pipeline capacity comes on line, but drilling activity jumps in all basins during 2h19 and 2020, on higher oil prices from global tightness and IMO 2020.  Typical cyclicality returns in 2022-23, as another LTO supply surge hits slipping demand from the 2020 price shock.

The other approach to stifling Permian production is to not complete wells.  This would further swell the inventory of drilled-but-uncompleted (DUC) wells in the basin, but this is probably the most effective approach.  As shown in the following chart, as of July 2018, EIA data indicates that Permian DUCs have surged by 80% yoy, to 3,468, which represents an inventory of 8.0 DUCs per well completion.  Producers will always have some inventory of DUCs awaiting completion crews, but this measure provides some indication of the backlog relative to completion activity, and hence, crews available.  Given the anticipated rigs, rig productivity, initial production (IP) rates and decline rates, Permian DUCs would need to rise to almost 5,000 by mid-2019, with DUCs per completion jumping to 14.8 in 1q19.  This would keep well completions near 400 per month and production limited to 3.6 mbpd, consistent with pipeline capacity until mid-2019.

The arrival of new pipeline capacity would prompt a release of the DUCs accumulated during the constrained output, but the anticipated rise in oil prices would also support a decline in DUC inventory.  The DUC cycle reverses once prices decline in 2022-23.

Although the EIA does forecast some flattening in the growth of US crude oil production in its Short-term Energy Outlook (STEO), our Base Case forecast is 200 kbpd lower during the last nine months of 2019, as shown in the chart below.

Our forecast then exceeds the EIA outlook by 900 kbpd during 2021, from higher oil price and resulting rig count assumptions.  For example, the latest STEO expects WTI prices of US$64.33/bbl during 2019, a 2.8% decline from their 2018 expectations of US$66.18/bbl.  One hint that the EIA price outlook may be too low — hardly any widening in the diesel-fuel oil spread in late-2019, as if IMO 2020 does not exist.

Oil market angst over recent jump in EIA crude production forecasts

In addition to recent financial markets turmoil, the crude markets have been under price pressure from concerns over surging US crude production, punctuated by a sharp upgrade in forecasts from the EIA in its Short-Term Energy Outlook (STEO) this week.  In fact, the February STEO featured a 317 kbpd jump in its 2018 forecast for US crude production from its January report, which was also 668 kbpd above its October forecast (shown above).

The April 2018 WTI contract has given up more than $5/bbl this week on these concerns, which may be overstated, since re-benchmarking and methodology adjustments have driven the surge in EIA estimates.  After all, the latest weekly estimate of US crude production (for the week ending 02 Feb) jumped by 332 kbpd week-over-week, to 10,251 kbpd, after averaging 28 kbpd of weekly gains during the previous 13 weeks.  Similarly, the monthly data for November 2017, released a week ago, surged by 384 kbpd over the October estimate, also reflecting the recent changes in EIA modelling and reporting.

The EIA actually tried to inform market participants about these changes ten days ago in a presentation, following up on a November 2017 webinar.  Most oil analysts recognise this and know that the EIA will make upward adjustments their historical database accordingly.  This higher level of crude production has already flowed through the system and the price movements reflect this.  The supply/demand balance has not changed and the near-term rises in production are consistent with the recent rise in rig counts and estimated rig productivity.  Only market perception has changed.

US Crude Production Jan16

US Production declines set to continue into 2017 — US light tight oil (LTO) production response is lagging rig counts by 6-7 months, based upon legacy well decline rates and rig productivity.  Our Base Case expects LTO rig counts to take another drop in 1q16 and remain low through most of 2016.  Initial signs of market re-balancing in 2h16 should allow a modest price recovery and some rig additions in late-2016.  LTO output needs to decline by 1 mbpd to balance global oil market, in addition to other non-OPEC production declines.  The December 2015 Short-term Energy Outlook (STEO) by EIA had forecast Brent prices at $56/bbl in 2q16, with resultant rebound in US production, but this is not realistic and does not balance the market.  Download presentation on US Crude Exports here.

US Condensate Production & Exports Jan16

Declining USG (PADD3) condensate production to limit export availability — Severe rig count reductions in the Eagle Ford play and accelerating well decline rates are sending Eagle Ford crude and condensate output much lower for 2016-17.   Although the repeal of the US crude export ban is further eroding the economics of condensate splitters on the USG, approximately 350 kbpd of new splitters are already completed, are under construction or have committed take-or-pay off-take agreements.  A combination of lower condensate production, rising splitter intake and Canadian diluent requirements should limit condensate cargo export availability through the forecast period.

 

US Crude Export Destinations Jan16

European refiners play larger role in US crude & condensate exports —  With Canadian imports remaining stable, Europe becomes the primary destination for incremental US crude and condensate exports.  Although North Sea production should decline by 300 kbpd during 2015-20, European refiners have limited ability to take larger volumes of 45° API crude and 55+° API condensate, when we examine each country’s crude slate.  US producers will need to find additional outlets in Latin America, as well as higher Asian exports, as the rising light-ends imbalance pressures prices and opens arbs.