Today’s report (week ended 29 Mar): EIA net change: +23 bcf, EA: +16 bcf
- Today’s reported injection was outside of the range of survey estimates and underscores how loose balances are currently. The market miss is likely hinting at lower-than-modelled res-com demand in combination with production (+1.3 bcf/d) growing at a faster clip than indicated by flow data. The massive injection into South Central inventories is also by far the largest ever for the region for March—by more than 1 bcf/d.
Next Thursday’s report (week ending 5 Apr): EA preliminary: +33 bcf
- Production growth throttles back to only 0.3 bcf/d w/w.
Waha’s weak week
Shoulder season is firmly underway. As expected, flush production following freeze-offs and maintenance at Sabine Pass T1 and T2 are setting the market up for lacklustre cash pricing. Our balances show production for the week in progress in line with peak weekly supply volumes registered in late December 2018, and Appalachia still has not recovered to its pre-freeze-off baseline. In other words, more sequential production growth is expected once Appalachian volumes fully recover. And the market is also seeing organic sequential production growth currently, not just a recovery. Flows are pegging y/y growth at 9.2 bcf/d, with about 1 bcf/d of the supply side being absorbed by a y/y decline in net Canadian trade.
The observed production growth is driving the market to be looser y/y by a whopping 7.5 bcf/d for the reference week and current week. However, this looseness reduces to just 2.5 bcf/d y/y if adjustments are made for the Sabine outage and assume similar weather y/y. Moving forward, the major question for the balances is whether the looseness can be reduced by new demand projects or exacerbated by further production growth. At least in the shoulder season, unless demand spikes on warmer-than-normal weather, balances are looking loose.
That is not to say that there are not sources of growth in our balances. Gas-fired power generation is up y/y to start April. As we set out in North America Outlook: Balances thaw, 29 March 2019, continued weak hydro output in the Northwest and displacement of higher-cost coal plants, are supporting higher y/y gas burn. In addition, nuclear outage season is deeper than expected so far. Several nuclear units in the east were shut for unplanned work in recent weeks, and another eight or so have yet to begin their expected spring refuelling, making what we had forecast to be a relatively light outage season something more significant.
However, ramp-up of Cameron feedgas has been back to zero in the past five days before today’s 5,000 mmbtu take, despite the facility receiving FERC’s blessing to commission its feed gas system on 27 March and to commission its dry flare system yesterday, although we expect some pickup in feedgas demand in the shoulder season. Incremental volumes heading into Corpus Christi T2 should also move the needle on demand before peak cooling season. However, realistically, another 1.5 bcf/d of demand from LNG alone could kick in this injection season as CC T2 and Cameron T1 ramp up, with incremental volumes from Freeport T1. No news on projects in Mexico suggests little to no movement in southern cross-border trade in the shoulder season. That said, if the Sur de Texas-Tuxpan pipeline does not miss CFE’s newest target by a wide margin, we estimate the market could see another 0.5 bcf/d of uplift.
How much this demand growth limits market looseness is still heavily dependent on how much organic growth we see in output as the market enters peak cooling season.
Waha: Any which way but tight
One region in which we do not anticipate fundamental tightening as the injection season progresses is Waha. For some time now we have underscored how periods of maintenance, especially in the shoulder season, are likely to result in negative pricing at the hub. The extreme negative pricing witnessed this week with some intraday transactions being conducted as low as -$9/mmbtu is indicative not only of how full all routes of the region are, but of how even a relatively minor amount of off-lined capacity can easily exacerbate congestion issues given how close to capacity the northern, western and eastern routes out of the region are.
The deep negative prices came as recent work on EPNG’s Florida compressor station finished on 31 March, restricting flows by 0.1 bcf/d. Northbound flows out of the basin have been limited by an ongoing force majeure on EPNG’s Line 2000, knocking off about 0.1 bcf/d of capacity. There has been word of maintenance on eastbound transport on the Oasis gas pipeline too.
Permian gas burn has been strong to date on a y/y basis, but seasonally we could see those volumes fall off as demand does (see Fig 1). The two CCGTs in the region- the 0.6 GW Hobbs in New Mexico and the 1 GW Odessa unit in Texas currently total near 0.2 bcf/d of demand. Even at full utilisation of these two plants and others in the region, the impact on demand growth is minor as they can only provide up to 0.4 bcf/d under peak cooling load conditions.
Our view on incremental relief from the Wahalajara pipeline is fairly negative and we think it is highly likely that delays could push back additions to the system till after the injection season (see, E-mail alert: Further delays to Wahalajara system highly likely, capping growth prospects for injection season US pipeline exports to Mexico, 19 March 2019). Meanwhile, we expect the next leg up in oil output to take hold in late Q2 19 based on new takeaway and frac crew increases.
For near-term Waha pricing this shoulder season, extreme basis weakness will be tied to maintenance and other operational issues. With work on EPNG’s Line 2000 ending late yesterday evening, some of the weakness in the past few days could be alleviated. However, that weakness is also attributed to maintenance on Oasis, with little confirmed clarity on extent or duration given it is an intrastate pipeline tying into other intrastate pipelines. A capacity reduction on Transwestern will reduce East Mainline receipt volumes by some 0.3 bcf/d through 18 April.
When producers are paying to take gas off their hands, value (or negative value) can get chased down. To put these values into perspective, we looked at a notional Permian well producing 60% oil, 30% NGLs and 10% gas. Even at super-negative pricing like yesterday’s -$5.79/mmbtu settle, given a 10% gas yield of the well, the weighted average gross revenue of the well would not be overwhelmingly impacted (see Fig 3). Using Parsley Energy as an example, unit costs per boe in Q4 18 were some $29. Looking at the economics for the full barrel, not just gas, and taking into account Opex, super-negative cash prices lower the bottom line but do not necessarily make the well uneconomic. Any fines levied for non-permitted flaring could hurt the bottom line more. However, total violations across all types levied by the Texas RRC last year totalled $9 million, so it may be the case that these fines are not a prohibitive obstacle for associated production.
|Fig 1: Permian basin CCGTs gas consumption, bcf/d||Fig 2: Projected EIA weekly storage change, bcf|
|Source: Ventyx, Energy Aspects||Source: EIA, Energy Aspects|
|Fig 3: Waha notional well weighted average gros revenue by gas price, $|
|Note: Assumed Midland oil price: $61.26/barrel, assumed NGL price: $34.00/barrel.
Source: ICE, Refinitiv, Energy Aspects