Blog April 02, 2019

Late on the evening of March 20, Genscape infrared monitors observed the first of two natural gas liquefaction (LNG) units, or trains, go offline at Cheniere’s Sabine Pass LNG facility. Our team observed these operational changes after scheduled pipeline delivery nominations headed to the facility indicated sharp declines over two consecutive days compared to normal gas volumes delivered to the facility. While initially presumed to be an unplanned maintenance issue, our team later attributed the outage to “opportunistic maintenance” as it coincided with a nosedive in gas prices in the Japanese (JKM) and European (NBP) markets over the last several weeks.

Historically, issues that result in a train shutdown often prompt either a maintenance event – that may or may not be publicized prior to the event – or a force majeure declaration. Shutting down one of these liquefaction trains could be the result of major, external market events or serious technical constraints at the facility itself, which could then ironically lead to additional fluctuations in the LNG market – or U.S. domestic natural gas market. For example, weather events caused numerous shutdowns at Sabine Pass in the past, such as during Hurricane Harvey or the intense fog this past February that prevented empty vessels from entering the port to load LNG. A high inventory level coupled with Sabine Pass’ inability to export volumes on ships caused the facility to stop producing LNG. While a combination of adverse weather and inventory constraints can cause serious issues at Gulf Coast liquefaction facilities, this was not the case for this current outage. Our LNG inventory models showed the facility was well below their inventory capacity, storing a mere 2.5 billion cubic feet (Bcf) of gas at the time the first train went down.

sabine pass scheduled pipeline deliveries in bcf/d
(Above) This graph depicts the scheduled pipeline deliveries to Sabine Pass LNG and illustrates when the fog-related outage occurred versus the current outage. 

Over the two days where Sabine Pass Train 1 and 2 eventually shut-down (March 21 and 22), we saw a 1.3 Bcf/d drop in delivery nominations compared to the prior two-week average. By late Saturday evening (March 23), both Trains 1 and 2 at Sabine Pass LNG facility stopped receiving gas and ceased all operations, as indicated by a lack of flare activity. This past winter, Sabine Pass LNG almost consistently remained 100 percent operational across all five of their trains. The facility operator boasts the capacity to liquefy more than 4 Bcf/d and has a strong reputation of maintaining continued service at trains.

sabine pass train ir image
Above is an infrared (IR) image of Sabine Pass LNG Train 1 (right) and Train 2 (left) fully operational on March 20 prior to the outage. The IR cameras detect heat signatures relative to the overall temperature reading across the viewing area. Liquefaction operations are indicated by the brighter heat signatures of the individual train stacks as well as the "white hot" exhaust plumes that top the stacks and the shared flare. 
sabine pass ir image of train 1 and 2 post outage
Above is an infrared image of Sabine Pass LNG Train 1 (right) and Train 2 (left) during the current outage event. Note that operations at the trains ceased and now appear "dark" with no exhaust plumes above the individual stacks. Also note that wet/dry flare shared by Trains 1 and 2 is not flaring, indicating a full shut-down of both trains. 

Rather than be solely reliant on pipeline nominations data, Genscape LNG subscribers benefitted from analyst insights informed by real-time monitoring of train operations. Our monitors were the only source of immediate insight as the trains shut down, because the facility operator did not offer any public announcement. More importantly, our real-time viewing of flaring activity – or lack thereof – provided a deeper understanding as to the expected duration of the outage. As Train 1 and 2 flares began to turn off, it was clear that a major maintenance event would commence at the two impacted trains. While this incident was most likely not originally scheduled to take place during the time, it’s plausible that the operator took advantage of the low foreign price/demand period and moved up the maintenance schedule.

Because Genscape monitors all train activity at Sabine Pass LNG – as it has since the inception of the facility – we can model the duration of outages. Our historical precedent shows Sabine Pass LNG Trains 1 and 2, and their related flare stack, have not undergone significant maintenance for more than two years. Additionally, in a recent earnings call, Cheniere alluded to a “higher than average” maintenance for summer 2019. Extended maintenance events, similar to the present outage, lasted roughly one month, however, analysts do not expect the current outage event to last that long. Currently, our team expects the outage to conclude within two to three weeks from the start of the outage (on March 20) with a few days of partial train startup, as the systems are tested post-maintenance.

Our analysts will continue to monitor the facility and will alert clients when monitors show the first signs of a restart. Real-time alerting and analysis of this outage is available via Genscape’s North American LNG Supply & Demand suite. The suite provides timely insight into train-level LNG production, storage, North American import and export volumes, and vessel tracking.