Handling LNG Feed Gas Variations By Jenny Zhang, Tariq Shukri and Tony Tarrant, Foster Wheeler

Abstract LNG liquefaction plants are designed to handle a range of feed gas compositions based upon the anticipated field production profiles. However, during detailed engineering or in operation the plant may be required to handle feed gas compositions which fall outside its design operating envelope. This paper discusses common issues arising from variations in LNG plant feed compositions that may occur during detailed design or in operation and presents some measures that can be taken to assess and accommodate the impact of such variations. The discussion is focused on the C3/MR process, although the same principles are applicable to other processes. The aim is to manage flexible plant operations through cost-effective measures, focusing on the modification of the flow and operating conditions, rather than immediately resorting to costly re-engineering or debottlenecking exercises.

Background Feed to LNG plants is composed primarily of methane, together with ethane, propane, butane and heavier components. Non-hydrocarbon components such as nitrogen, carbon dioxide, hydrogen sulphide and mercury are also usually present. A typical range of feed composition is shown in the table below. Methane

CH4

Ethane

C2H6

Propane

C3H8

Butane+

C4H10 +

Carbon Dioxide Nitrogen

70-90%

0-20%

CO2

0-8%

N2

0-9%

Table 1: Typical LNG Feed Gas Composition [1]

During the early design stages of a project, the range of feed gas composition is established based on the field data available at that time. Engineering judgement is used to review the compositions and production scenarios to fix the design envelope for the plant. This leads to a clear design basis for the front-end engineering design with a limited number of design compositions that dictate equipment sizing. At this point, there may be a temptation to specify a very wide range that will accommodate every possible composition, but this will lead to large design margins and rapidly escalating costs. The skilful practitioner will be pragmatic in specifying the range of compositions, knowing that he may be able to utilise the inherent margin at a later stage if changes have to be made.

It is not unusual for the predicted feed gas composition to vary to points outside the range specified in the initial design basis. Improved data may become available during the detailed engineering or new gas sources become available during operation. An experienced design contractor will know how to manipulate the existing design, pushing the boundaries of operation so that production may be maintained. This paper addresses some common LNG plant challenges resulting from feed variations and outlines cost-effective design and operational options to maintain LNG production and specifications.

Variations in the LNG Feed Gas By design, plants can handle different feed gas compositions. For short-term changes, such as interruption to upstream facilities, a degree of impact on production may be acceptable so long as product quality can be maintained. However for longer term operation, measures may have to be taken to meet production requirements without compromising on quality. The operating envelope of an LNG plant is typically defined in terms of its average, lean and rich feed gas compositions, as shown in the example in Table 2 below. It is necessary for the designer to understand the key plant design parameters and their impacts on the plant operating flexibility in order to ensure the robustness of the design and operation, and to maximise investment returns.

LNG Feed Nitrogen CO2 Methane Ethane Propane Butane Pentane C6+ Total

Average Gas Mole % 5% 2.2% 85% 4.5% 1.7% 0.8% 0.4% 0.4% 100%

Lean Gas Mole % 7% 1.8% 87% 2.6% 0.9% 0.4% 0.2% 0.1% 100%

Rich Gas Mole % 3% 1.9% 83% 7% 2.2% 1.4% 0.6% 0.9% 100%

Table 2: Example LNG Feed Gas Composition [1]

In the industry, a “rich” feed means a feed gas containing a higher amount of hydrocarbon components heavier than methane (i.e. ethane, propane, etc.), whilst a “lean” gas contains smaller proportions of these heavier hydrocarbons or a higher content of inert gas, such as nitrogen. For LNG production, the feed gas typically requires treatment to reduce constituents such as water, carbon dioxide, nitrogen, oxygen, mercury and sulphur compounds.

Fractionation and

natural gas liquid (NGL) extraction may be required to recover hydrocarbon components for refrigerant make-up and for LPG and condensate product streams.

Figure 1 below shows a typical block flow diagram of an LNG production plant and Figure 2 shows an outline process flow diagram for the liquefaction section of a typical propane pre-cooled mixed refrigerant (C3/MR) plant.

Figure 1: LNG Block Flow Diagram [2]

Figure 2: Typical C3/MR LNG Process [2]

Performance Assessment The impact of a change in feed gas composition is initially assessed by performing process simulations and hydraulic calculations. The first aim is to assess performance of the process configuration as designed when operated with the new composition. If this indicates a shortfall in the nameplate capacity or a failure to meet product specifications, then the next phase is to assess how the performance can be enhanced to meet production targets, identifying bottlenecks and operational changes needed. The design / performance of the major equipment in the process is primarily used as a constraint to investigate if changes to operating parameters can be accommodated.

Various modes of

operation have to be considered including normal, start-up and shutdown. Examples of typical major constraints are: a.

Refrigeration compressors

b.

Compressor drivers

c.

Main cryogenic heat exchanger

d.

Hydraulic limits of equipment and pipework.

Any proposed changes need to be considered holistically, ensuring that: Safety systems are not compromised; The operation is sustainable with respect to utility and support systems, such as refrigerant make-up; The LNG product conforms to the required specifications, including the following, which are relevant to subsequent discussion: 

Maximum allowable amount of heavy hydrocarbons, typically 0.1 mol% C5+ and 2 mol% C4+.



Maximum allowable N2 content, typically 1 mol% in loaded LNG.



Specified range of calorific value and Wobbe Index.

Measures to Address Changes In the following sections, a number of design and operational measures are described which can be implemented to mitigate feed gas changes without major modifications to process plant. For the purposes of this article, they mainly relate to operation around the scrub/heavy hydrocarbon (HHC) removal column, although the impacts will likely extend into NGL fractionation and the upstream gas and condensate processing. The built-in flexibility of the process design arising from the specification of a range of feed compositions, and the margins on equipment design imposed by both the engineering contractor and suppliers, can be utilised to accommodate changes to the feed by manipulating operating conditions and installing minor modifications.

Richer Feed Gas Generally speaking, a rich feed containing a relatively large amount of C2+ is good news for the operator, since higher production rates may be achievable. A richer feed gas results in a heavy loading of the liquid processing units of the LNG plant, including NGL fractionation and condensate stabilisation. For a rich gas case where the required liquid processing load is higher than the design loading, the LNG train would have to be operated at turndown to avoid overloading the liquid processing units. To restore LNG production, these units would have to be unloaded or debottlenecked. CONDENSATE PROCESSING Front-end condensate processing equipment normally has a large excess capacity margin designed to handle gas trunkline liquid slugs, and so should therefore be able to process increased condensate production in normal operation. Increased liquid content within the pipeline may result in more frequent arrival of slugs or larger slugs during ramp-up periods, which require more focused operation of the slug catcher. Effective control of the slug catcher liquid levels and buffer storage capacity within condensate storage facilities can usually accommodate increased rates, although it may be necessary to tolerate reduced liquid residence times. NGL FRACTIONATION After acid gas removal and dehydration, the heavier hydrocarbon components are removed in the scrub column and further separated in the NGL fractionation unit. C5+ components form the bulk of the liquid from the bottom of the last fractionation column, which is mixed with any front-end stabilised liquid to form the condensate product. Other NGL components from the overhead of the fractionation columns are either used as refrigerant make-up components, exported as LPG products or re-injected back into the main natural gas feed upstream of the main cryogenic heat exchanger (MCHE). Opportunities exist for debottlenecking the NGL facilities, without necessarily resorting to expensive internal replacement exercises: this is outside the scope of this paper. This article considers how the fractionation unit may be unloaded through operation of the scrub column. This can be achieved by modifying the column operation and/or bypassing some of the reflux directly for re-injection into the feed to the liquefaction exchanger. 

Scrub Column Reflux Partial Bypassing to MCHE

To unload the scrub column and fractionation units, the scrub column reflux drum liquid can be split, with one part directed forward and re-injected upstream of the MCHE. This partial bypass reduces column reflux as well as the liquid processing load to the fractionation unit, as shown in Figure 3 below. The split ratio in the chart is defined as the ratio of the bypassed reflux rate to the total liquid rate from the overhead condenser.

Although the split may help to unload the fractionation and the scrub column system and therefore debottleneck the system, it is limited by two constraints: 1.

The impact on the LNG product calorific value, i.e. whether the product’s higher heating value (HHV) exceeds the maximum specification. Depending on the existing plant flow scheme, if LPG is designed to be exported as a product, the HHV of the LNG product will increase with increased reflux bypass.

On the other hand, the impact will be

minimal if there is no LPG product and all of the LPG is re-injected. 2.

The impact on the scrub column separation, especially on the C5+ specification. The amount of scrub column overhead C5+ has to remain within certain limits to avoid freezing in the MCHE and also to meet the LNG product specification.

Flow to Fractionation C5+ Mole Fraction

0

0.1

0.2

0.3

0.4

Split Reflux Ratio Figure 3: Scrub Column Reflux Partial Bypassing to MCHE [1]



Scrub Column Feed Temperature

The scrub column feed temperature influences the separation and the heat balance of the column. Increasing the feed temperature will increase the flash of the heavier components and reduce the bottom liquid rate. To unload the fractionation unit, the scrub column feed temperature can be manipulated to reduce the column bottom flow to the fractionation unit. This temperature is important for the column separation, especially for the separation of heavy hydrocarbon components from the natural gas stream.



Scrub Column Overhead Condenser Temperature

One option to debottleneck the fractionation system is to manipulate the scrub column overhead condensing temperature by bypassing part of the overhead cooling duty. This reduces the amount of reflux to the column and hence the temperature profile and column bottoms flow to the NGL train, as shown in Figure 4, below. Again, careful evaluation is required to ensure optimum operation within constraints, such as the required scrub column separation and to prevent freezing of heavier hydrocarbon components in the MCHE.

Reflux Flow Flow to Fractionation

-65

-60

-55

-50

-45

-40

-35

-30

-25

-20

Reflux Temperature, °C Figure 4: Impact of Scrub Column Condensing Temperature [1]

Leaner Feed Gas The operator that finds itself with a lean feed gas containing more inerts and less C2+ than designed for, is likely to face more problems in trying to achieve the nameplate capacity. High nitrogen content is particularly restrictive as it absorbs refrigeration capacity without providing any additional product. System capacity limiting factors typically exist in the following areas: REFRIGERATION POWER Leaner gas requires higher specific power. production targets.

Hence, more power may be needed to meet

Options to uprate compressor drivers and helpers may be considered.

Analysis on the refrigeration power is an integral part of LNG production optimisation which forms its own territory, outside the scope of this article.

PRODUCT CALORIFIC VALUE Lean feed gas often leads to low heating value of LNG product which may not meet the product minimum specified higher heating value (HHV). The theoretical maximum achievable product HHV for a particular lean feed can be evaluated based on the feed composition excluding inert gases and its condensate contents.

If this

theoretical maximum value is still below the required specification, it can be concluded that the LNG product minimum calorific value will not be met independently on site. Additional material such as LPG components may have to be injected into the LNG at the source or destination to raise its HHV. REFRIGERANT MAKE-UP Ethane and propane refrigerant make-up is required for start-up and to compensate for any refrigerant loss in normal operation such as losses in the compressor seal system. The design make-up rate is usually set by the need to fill the refrigerant storage within a reasonable amount of time. Storing spare inventory helps to secure plant operation in case of serious upsets. With low ethane and propane content in a lean gas feed, there may be insufficient recovery in the fractionation unit to meet the refrigerant make-up requirement. A number of operational changes may be considered to increase the recovery of refrigerant components, whilst trying to minimise the impact on the existing design. 

Scrub Column Reflux Bypassing to Fractionation

Part of the liquid from the scrub column reflux drum can be comingled with the column bottoms to increase the rate of refrigerant for recovery in the fractionation unit, as shown in Figure 5. The reflux bypass ratio here refers to the amount of the liquid bypassed to the column bottom divided by the total amount of the liquid generated in the overhead condenser. Reflux bypassing in this manner will increase the liquid load to the fractionation unit and help to boost ethane and propane recovery to meet the make-up demand. However, care must be taken to ensure that sufficient reflux is still available for the column to remove C5+ components effectively.

The column pressure can be reduced to enhance component

separation, as this moves the column away from the critical region. Careful analysis is required here to consider the subsequent impact on liquefaction efficiency and available power.

Ethane flow to Fractionation Unit Propane flow to Fractionation Unit 0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

Reflux Bypass Ratio Figure 5: Partial Scrub Column Reflux Bypass to Fractionation [1]



Scrub Column Reflux Temperature

It may be possible to operate at a lower temperature in the scrub column reflux drum to increase the reflux rate and boost the recovery of ethane and propane from the fractionation unit, as shown in Figure 6 below.

Ethane flow to Fractionation Unit Propane flow to Fractionation Unit

-66

-64

-62

-60

-58

-56

-54

-52

-50

-48

-46

-44

-42

Temperature, °C Figure 6: Impact of Scrub Column Condensing Temperature [1]

-40

As with the use of a reflux bypass, there are knock-on impacts from reducing the condensing temperature which must be examined: a)

The reduction is possible only if there is sufficient cooling duty in the system to remove the heat, and this can lead to increased demand on both propane and mixed refrigerant circuits.

b)

The condensing temperature should be determined together with the operating pressure, so that operation can be maintained well away from the critical point to ensure vapour liquid separation within the reflux drum.

Conclusions Moderate changes in feed gas composition may be accommodated without significant modifications to key equipment through the adjustment of operational parameters or addition of partial bypasses, which can be allowed for during detailed design. Sufficient flexibility is often available as a result of designing the process for a range of compositions and the inherent design margins in individual pieces of equipment and piping. Through insightful analysis of the process, these margins can usually be squeezed to bring the plant back to its desired capacity. However, this requires comprehensive knowledge of the overall process and the interactions between units. Without this understanding, an apparently simple change to one process parameter may throw other sections of plant out of balance.

© 2012 Foster Wheeler. All rights reserved.

AUTHORS’ BIOS JENNY ZHANG Jenny Zhang is a principal process engineer working as a member of the process technology group within Foster Wheeler's process engineering department in Reading, UK. She has over 19 years' experience, specialising in process design, simulation, process RAM studies and energy optimisation covering refining, LNG, GTL and CCS projects. Zhang holds a MPhil degree in chemical engineering from UMIST, Manchester, UK, and a BSc in chemical engineering from East China University of Science and Technology, Shanghai, China. Contact: [email protected]

TARIQ SHUKRI Tariq Shukri is a chemical engineer with extensive experience in the operation and design of LNG and hydrocarbon processing plants. He joined Foster Wheeler in 1995 after a long spell in industry providing process engineering and technical support for the operation and design of LNG and gas

processing plants. His current role is chief engineer - LNG for Foster Wheeler's operation in Reading, UK, where he provides technical expertise to LNG feasibility studies, FEEDs and EPC design projects. Tariq holds an MSc and PhD in chemical engineering, both awarded by the University of Newcastle upon Tyne, UK. Contact: [email protected]

TONY TARRANT Tony Tarrant has over 18 years' experience in process design, primarily focused on offshore production, onshore gas processing and LNG. He is technical manager for Foster Wheeler's Business Solutions Group, which provides high-value consultancy services to operating companies through early project phases from feasibility studies, conceptual design to pre-FEED in the oil & gas, midstream (LNG, XTL and CCS), downstream (refining and petrochemicals) and speciality products areas. Tony is a chartered chemical engineer and holds a BEng in chemical process engineering from the University of Aston in Birmingham, UK. Contact: [email protected] REFERENCE 1. Foster Wheeler in-house data. 2. Tariq Shukri, “LNG Technology Selection”, Hydrocarbon Engineering, February 2004.

NB: An edited version of this paper entitled “Foster Wheeler outlines its latest process for handling feed-gas for the production of LNG” was published in the May 2012 edition of LNG Journal.