Part 1 of this series (which appeared in the July/August 2013 edition) provided a historical review of LNG turbomachinery trends. It showed that each LNG project has its own unique set of upstream costs, layout considerations, operational factors and governing economics. In addition, it highlighted the fact that economies of scale have flattened out and that cost savings are less likely to be found in “going bigger” in terms of turbomachinery train configurations.

Part 2 assesses turbomachinery and process cycle selection for LNG projects. In evaluating turbomachinery in the context of the overall plant design, the project team must consider factors such as machinery degradation rates, operator experience, spare parts availability, expansion planning for additional trains and the machinery emissions signature. However, to shorten the present discussion, the two primary factors influencing LNG project net present value (NPV) are discussed here: specific energy consumption and availability.

The conclusion is that availability is more important to NPV than energy consumption. Availability will influence LNG project NPV more drastically than small percentage changes in efficiency or the process cycle efficiency. In an environment of LNG plants with higher capital cost, therefore, validation of compression train drivers or improvements to the refrigeration cycle are difficult to justify given the risk to NPV if plant availability suffers.

Thermal efficiency

Thermal efficiency is often regarded as one of the most important decision criteria in deciding between competing refrigeration cycles and turbomachinery. However, the critical issue for LNG plants is specific energy consumption (SEC) or the amount of energy expended to produce LNG, which includes all of the plant processes (not only the major equipment). This metric also provides an indication of fuel consumption and process efficiency.

SEC is a preferred efficiency metric for an LNG plant as it accounts for all of the plant processing and machinery efficiencies; it can be measured in real time; and it is normalized such that different LNG processes or plant sizes can be compared fairly. SEC is the amount of energy consumed by the plant (feed gas heating value and electricity used), minus the amount of LNG and extracted Natural Gas Liquids (NGLs) delivered by the plant, divided by the energy consumed:

Energy consumption can be calculated for either the “holding mode” or the “loading mode” of the plant.The energy consumption will differ slightly because of the greater energy expended by the loading pumps when LNG is being transferred from the tanks to the carrier. This SEC definition provides the opportunity to refine the processing horsepower, optimize NGL extraction, add waste heat recovery, and improve the turbomachinery efficiency in order to compare different plant designs in terms of specific energy consumption.

It should be stressed that other decision criteria, specifically availability and operational flexibility, may be equally as important to plant economics as the plant’s energy consumption or process efficiency. In some cases, where supply gas is plentiful and relatively low cost to produce, $1.0-$1.50/ MMBTU, the plant’s energy consumption is less meaningful to the plant NPV than maintaining high availability.

The fuel costs are a project specific variable which affects how strongly the LNG plant fuel consumption is valued and the production costs. In an integrated upstream project, it is difficult to attribute appreciable gains in NPV to gas feed costs. It is important to not restrict the plant design methodology to simply the minimization of power consumption. As noted by Amin Almasi in the 2013 Turbomachinery International Handbook, “the resulting LNG plant could end up with a poor driver, high initial costs or an overly complex LNG and power plant system.”

Plant SEC also varies with ambient temperature, load and feed gas composition. An LNG plant’s SEC should be viewed as an operational range of achievable plant energy consumptions instead of a single “high efficiency” point, which misrepresents the true fuel usage variations on a day-to-day basis.

Above 25 MW for simple cycles, aeroderivative engines tend to have slightly higher maximum thermal efficiency than comparably sized industrial turbines (42% vs 36%, respectively) but this is model specific. However, aeroderivatives are typically not optimized for combined cycle and tend to produce overall lower combined cycle efficiency than industrial Frame 5 or Frame 7 gas turbines because of the higher exhaust temperatures of the industrial turbines.

In addition, thermal efficiency should be evaluated over the entire range of the foreseeable plant load. As indicated in Figure 1, each turbine and compressor system is different and will produce a different load curve such that some machines will be more sensitive to load reductions with greater declines in overall turbine and compressor system efficiency.

These same curves can be drawn for the change in thermal efficiency as a function of ambient temperature. For LNG plants, the load tends to be relatively constant, above 90%, which is close to the maximum design efficiency but the ambient temperature can produce swings of up to four efficiency points.

Compressor design improvements in the last 20 years have resulted in polytropic efficiency gains, improving the efficiency from 65% to above 80%. Some of these gains have been aided by advanced CFD to reduce pressure drop and improve mixing of the side streams.

Other innovations, such as oil-free magnetic bearings, have helped to reduce operational expenses. This 10%-plus improvement equates directly to lower energy consumption. Improving fuel consumption through driver and compressor efficiency and incorporation of combined cycle systems has significantly reduced energy consumption of LNG plants.

Efficiency opportunities

Process optimization, cycle selection and heat exchanger improvements have significantly less impact on plant energy consumption than gains made through machinery design, because the efficiency opportunity gains in the equipment are so much larger. This issue is enumerated well by the study put together by ConocoPhillips.

Another feature of modern plants is the incorporation of waste heat recovery to some extent at the turbine exhaust. To reduce heating power draw, waste heat systems to heat oil or water are often considered. However, a full combined cycle implementation has been accomplished at Tangguh LNG which uses steam for power generation.

The addition of a combined cycle system may lower the availability by 1 to 3 percentage points, but it can certainly boost the overall power cycle efficiency to more than 50%, especially when using an industrial gas turbine with higher exhaust temperatures. A carbon tax structure will provide greater cost justification for waste heat recovery systems. However, the availability penalty and additional capital cost of combined cycle must be seriously weighed against potential gains in efficiency.

Another factor to consider is that electric motors compete fairly well with gas turbine drives in terms of efficiency. But SEC is less important than the cost of power as electrical power cost must be compared directly to the cost of gas fuel.

In terms of availability, electric motors do not require inspection and overhaul periods so motors can offer higher initial availability compared to gas turbines. However, the reliability of electric motors must be ensured through a well-designed VFD motor compressor system. Torsional harmonics and inter-harmonics within the operating speed range can pose a problem to the entire train reliability.


Availability is more important to the NPV analysis than plant energy consumption or thermal efficiency. Plant availability can vary dramatically based on spare parts availability, maintenance intervals, operator decisions and plant flexibility. Unfortunately, since this metric tends to be operator specific and is best tested over time, availability metrics between LNG plant designs are hard to obtain.

In comparing aeroderivative and industrial gas turbines, it should be noted that the actual operator maintenance management and scheduling are the largest influences on availability curves over time. Aeroderivative engines are easier to overhaul due to their lower weight and size (2 to 3 days downtime).

One strategy often employed to minimize downtime is to purchase a spare engine or participate in a joint company spare engine program. The aeroderivative engine module can be replaced within a couple of days if a spare engine is available. Industrial turbines, on the other hand, will require longer overhaul times but the overhaul periods are less frequent.

In general, aeroderivatives tend to have more unplanned downtime and can have more steep degradation rates due to their basic design (thinner blades, less tip clearance, less robust). Incorporating the same type of gas turbine into each additional train design and in the same train design will help to ensure the same maintenance regime and spare parts catalogue.

A good example of this practice is the Atlantic LNG train expansions which have selected the same gas turbine model (Frame 5D) for its Trains 2, 3, and 4 over the years. The operators were able to increase train capacity and reduce energy consumption using the same gas turbine drives and process cycle.

Another system factor which can influence the availability is the degree of power augmentation or steam generation, both of which introduce more pieces of equipment and auxiliary systems which can negatively influence availability. Aeroderivatives will benefit more readily from inlet air cooling as a means of power augmentation in warmer climates.

Industrial units have higher exhaust temperatures and may include combined cycle operation to reduce the plant specific energy consumption. Combined cycle systems require a 2x to 3x larger footprint than a simple cycle plant.

An integrated plant power system will also be required in addition to equipment for steam generation, large heat recovery steam generation units, water treatment and watersteam handling systems. One can argue that this additional equipment and system integration will reduce availability by 2 to 4 percentage points.

Nonetheless, most combined cycle plants allow for simple cycle operation in the event that the steam generation equipment is down such that overall LNG production and plant availability is only mildly affected.

Net Present Value

An economic analysis was conducted to illustrate the effects on NPV due to different gas turbine driver efficiencies, simple vs. combined cycle operation and variations in plant availability. The six cases in Table 1 were compared.

In all cases, the train capacity was held constant at 5.5 mmtpa, a single train facility was assumed and total energy consumption estimated at 305 kw-hr/ton of LNG. The analysis also varied the gas feed price to show the changes to NPV based on gas feed with a constant LNG price of $13/MMBTU. Other assumptions were required, based on typical gas turbine efficiencies and availabilities in the technical literature as follows:

• The baseline case for all NPV comparisons was an industrial gas turbine drive operating in combined cycle mode, with the stated train capacity and plant energy consumption. Specific energy consumption in the base case was 8.25%

• For the capex basis, the simple cycle system was assumed to cost $1,200 per ton and the combined cycle system was assumed to cost $1,300 per ton

• The NPV was calculated on actual cash flow, for a 20-year period with a 10% discount factor • Energy consumption varied with simple or combined cycle and GT efficiency, all cases assumed equal compressor efficiency. Calculated values are shown in Table 1

• Availability assumptions: Industrial GT in simple cycle = 98%, Aero GT in simple cycle = 97%, All combined cycle systems have 2 percentage point deduction

• The industrial turbine declining availability curve (Figure 2) cases C-1 and C-2 is an estimation of a poorly executed maintenance program where degradation of the driver affects the overall plant availability, to illustrate the effects on project NPV

• No additional economic credit was awarded in the NPV analysis for efficiency gains or reduced carbon footprint (no carbon taxation).

Based on these assumptions, Table 1 summarizes the case studies which were considered in the NPV analysis. These were selected to compare the efficiency and availability differences in industrial and aeroderivative gas turbines, as well as the use of combined cycle to improve the efficiency of the plant.

To illustrate the importance of maintaining high availability, two cases considered a declining availability curve over a 20-year operational period. This curve is a “worst case” description and is based on severe gas turbine degradation rates. Figure 2 shows the availability decline curve which was used in Cases C-1 (combined cycle) and C-2 (simple cycle). These declines are based on gas turbine technical papers investigating filtration systems (or lack thereof), and online and offline washing programs for various gas turbine models.

The net present value of the baseline case plant (Case A-1) was compared to the NPV of the other five plant configurations. The difference in NPV from the base case is shown for each plant case as a function of gas feed price in Figure 3. The industrial and aeroderivative turbine operating in simple cycle mode result in approximately the same project NPV (within $25MM of each other) and both cases have a higher NPV (positive delta NPV) than the base case, which includes combined cycle.

This trend is indicative of the higher capital cost of the combined cycle plant and the non-realized economic benefit of the added efficiency, in a non-carbon taxation environment.

Comparing the combined cycle plants directly shows that an aeroderivative gas turbine operating at 97% availability results in a slightly less NPV than the base case with an industrial turbine operating at 98% availability. This comparison shows that a 1% difference in availability affects NPV more than the 6 point difference in GT thermal efficiency.

The combined cycle cases are more sensitive to availability variations because of the increased capital cost of the plant (+$100/ton was assumed). Finally, the two cases of declining availability illustrate the severe effect on NPV caused by steep degradation rates and resulting declines in plant availability. The two curves for C-1 and C-2 also illustrate the greater influence of plant availability on NPV when higher capital costs are incurred due to the incorporation of a combined cycle system.

According to this analysis with the assumptions provided, availability is more important to the NPV analysis than plant specific energy consumption, all other factors being equal. Plant availability is a function of the basic design, equipment selection and operator decisions on issues such as spare parts management and maintenance planning.

Penalties and cost

Availability will affect the LNG project NPV more drastically than small percentage changes in efficiency or the process cycle efficiency. The potential availability penalty and additional capital cost of combined cycle must be anticipated and managed against the potential gains in efficiency.

In a cost environment of higher capital cost LNG plants, validation of compression train drivers or improvements to the refrigeration cycle are difficult to cost justify given the risk to NPV if plant availability suffers. Efficiency should be measured by the plant’s specific energy consumption. This metric fairly accounts for all energy efficiency gains and can be used to compare plant efficiency improvements in terms of overall energy expended to produce LNG.


Dr. Klaus Brun is the Director of the Machinery Program at Southwest Research Institute. Klaus.brun@swri.org