LNG and digitalisation [Gas in Transition]

Go back only a few years and the LNG market was in a very different place. With low prices and a surfeit of new projects on the board relative to demand, an acceptable supply-demand gap that warranted a final investment decision on a new plant seemed to be permanently receding.

To succeed, projects had to show that they could produce LNG at significantly lower cost, both in terms of capital and operational spending, than the competition provided by existing plants and those on the drawing boards of competitors. They also then had to demonstrate their capacity to reduce greenhouse gas (GHG) emissions. Attracting long-term customers depended not just on cost, but the environmental impact of liquefaction.

The harnessing of today’s computer processing power and digital tools played an important role in the development of new concepts, for example modular mid-scale LNG plants, as opposed to pursuing ever larger economies of scale, with ever larger trains, which were reaching their practical, physical limits. Being able to simulate with digital models the myriad possible configurations ahead of construction, both for cost and GHG emissions, was an invaluable part of developing new low-cost, environmentally-acceptable LNG concepts.

Operational performance

Today, as Russia’s invasion of Ukraine has sent European demand for LNG skyward, cost is less of an immediate issue. Operational efficiency and reliability have come to the fore. US LNG plants, and others around the world, are operating at peak load to meet demand in a price environment that is paying rich dividends for those that can achieve sustained high levels of production.

But peak load operation is by definition not sustainable. Running equipment hard increases wear and tear and all plants need routine maintenance both for operational and safety reasons.

Moreover, if something does go wrong, the cost of damage to an LNG plant can be significant.

The Hammerfest LNG plant in Norway, for example, only resumed operations on June 2 this year, having been out of action since September 2020, owing to a fire which occurred in the filter housing on one of the plant’s gas generators. In this incident, in which there were thankfully no fatalities, the company identified as the cause the spontaneous ignition of the filters in the turbine’s air inlets, owing to excessively high temperatures over a long period.

More recently, in June, an explosion at Freeport LNG forced the US’s second largest LNG plant offline at a time when Europe is desperate for increased LNG supplies to reduce its dependence on Russian pipeline gas.

Partial liquefaction operations at the 15.3mn mt/yr, three-train Freeport LNG facility are not now expected until early October with year-end targeted for a return to full production. The explosion, which also caused no injuries, occurred in pipe racks that support the transfer of LNG from the storage tanks to the terminal’s dock area. None of the major site equipment – liquefaction trains, storage tanks, docks or LNG process area – were damaged, according to the company. Yet, Freeport LNG will still suffer significant downtime at a time of buoyant market demand and high prices.

The incident will be investigated thoroughly to determine the exact cause, but, in a general sense, the fact that something happened, somewhere in the LNG sector, was almost inevitable as LNG plant owners run their equipment hard for a relatively long period to meet customer demand.

Forewarned is forearmed

This is where the benefits of digitalisation again come to the fore, especially when set against the potential physical damage and lost operational time of a significant incident.

Proactive and predictive monitoring is central to the safe and reliable use of critical equipment and it begins with the design phase. Sufficient and effective sensing and monitoring equipment needs to be installed to collect the operational data necessary to evaluate not just the current performance of equipment, but to detect even minor changes that can predict potential failures ahead of time.

Predictive maintenance uses a data-driven approach to assess the state of field equipment or infrastructure, developing a forward-looking picture of its operational lifetime. Overall, this facilitates a fundamental shift from a reactive to a preventative approach to maintenance, which has a substantial impact on safety performance.

Data analysis can be enhanced by the creation of a digital twin of the whole system, or part of it, to simulate operational modes and their impact on equipment failure under well-defined operating conditions. Monitoring equipment also plays a central role in providing data for optimisation and improvement. The result is often a double win in terms of reduced maintenance costs and increased reliability.

In this sense, digitalisation – a rather all-encompassing concept – in practice means the gathering, processing and then use of much larger quantities of data than were previously available.

Each step poses its own challenges. It is one thing to collect data, but another to understand and make good use of the data collected. This is where computer power comes in again, assessing and analysing large volumes of data to identify, for example, deviations from predicted performance.

Wider value chain

The use of digital platforms that aggregate data allows more effective information sharing, which helps to break down the information ‘silos’ which have bedevilled the oil and gas industry in the past. In turn, this makes possible more integrated asset management across the entire LNG supply chain from upstream gas production through liquefaction to transport and regasification.

The number of sensors installed in LNG carriers, for example, has grown steadily, so that the ships being built today have a constant stream of operational data flowing to the central computer system and to data centres on land.

This allows remote fault diagnostics and maintenance decisions, either immediate or scheduled for the next suitable opportunity, for example at the next port, where needed parts can be delivered ahead of time and installed on arrival, reducing maintenance time.

A recent deal between Finland's Wartsila and NYK LNG Ship Management provides a good example of how digitalisation is improving predictive maintenance in LNG carriers. The 15-year deal aims to optimise all maintenance procedures to maximise operational availability and ensure long-term maintenance cost predictability, using artificial intelligence (AI) and advanced diagnostics.

Today’s LNG market demands sustained high production performance alongside stringent safety standards from all parts of the value chain. Whether for a gas compressor in an LNG plant or an LNG carrier, digital modelling and diagnostics play an increasingly important role in hitting and maintaining that operational sweet spot, which safely combines high levels of production and performance with low maintenance costs and downtime.