Article %2d Foster Wheeler Energy%2epdf

ENGINEERING

SOLUTIONS

Picture Courtesy: ConocoPhillips Company

Engineering A Gas-to-Liquids Project

TAKING GTL FORWARD Gas-to-Liquids (GTL) has been the subject of many press conferences and journal articles in the last four to five years. Attention has been focused on catalyst developments or improvements, the undoubted outstanding quality of the product fuels, and the challenges facing us as engineers. Most people are aware of the basic economics behind GTL and the fact they are not clear-cut for most regions of the world, even though, on paper, typical reported gas prices correspond to feedstock prices of around US$ 4-6/bbl. However, we must now realise that the time for these type of debates is over. GTL projects have come of age and the engineering challenges that have been debated at length over recent years have been resolved. Arguably, the economic arguments have also been overcome for certain regions of the world, and we are seeing projects in these regions on the verge of construction. This paper will, therefore, concentrate not on debate and challenge, but on real engineering solutions that have been developed for a range of GTL projects. We will also attempt to draw some conclusions for the next-generation facilities now being planned for later in the decade, which will build upon the experiences of these first plants. SIMON C CLARKE & BAHRAM GHAEMMAGHAMI

OFFSHORE WORLD OCT-DEC 2003 n 55

ENGINEERING

SOLUTIONS

T

he three main process technologies in a Gas-to-Liquids (GTL) process are well known: l Syngas preparation l Fischer-Tropsch (FT) synthesis l Hydrocarbon upgrading (mild hydrocracking/hydrotreating). Much has been written about the advantages and disadvantages of a variety of proprietary technologies for each of these steps. Debates have raged over partial oxidation, catalytic partial oxidation, autothermal reforming, slurry beds, fixed beds and so on. However, all of these debates still do tackle the area, which accounts for 40 to 50 per cent of the cost – the utility and offsite support systems. Virtually all the technologies in a GTL plant have a common utility thread: l The need for large quantities of high grade energy to drive the air separation processes l Preheat needs for the syngas generation step l Waste heat recovery from syngas and its effective utilisation

l

l

l

l

Tankage and product-loading facilities, which, when compared to a refinery, may require similar volumes, but whose utilisation can be low (when considering projects of around 34,000 barrels per day (bpd)) Systems that require virtually 100 per cent sulphur quarantine to protect sulphur sensitive FT catalyst (tankage, flare and vent systems) Large-scale, reliable electrical systems, often internally generated derived from waste heat Usual support infrastructure of administration buildings, workshops, warehouses, canteens and medical facilities

l

Supported by "power station"-sized steam and electrical systems, a wastewater treatment facility, and associated infrastructure.

Process Engineering Opportunities

The challenges listed above are in fact unique opportunities within a GTL facility that can be exploited as a benefit. GTL projects are unlike any other, with the particular combination of facilities within a single project entity totally different from any other refinery, chemical or powerbased project. Rather than being a hindrance to commercialisation, it should be looked at as a differentiator.

Figure 1: Basic GTL flow scheme. Medium/low grade heat generation by the FT process l Hydrogen provision for the hydrocracker l Optimum product recovery to maximise yield l How to economically reject ~40 per cent of the feed natural gas heat (GTL projects are around 60 per cent thermal efficiency, resulting , therefore, in around 40 per cent heat rejection to the surroundings). In addition to this, the offsite systems are also significant, especially when dealing with greenfield remote locations: l Water treatment to reliably support the large steam systems l Effluent treatment of oxygenated hydrocarbon contaminated water (and utility system blowdowns) l Flare systems, dealing with high heat flows from the hydrocarbon processing units and high hydraulic flows from the gas processing units l Firefighting systems, dealing potentially with large volumes of hydrocarbons at their boiling points and hydrogen containing streams l

Large-scale temporary construction facilities; a significant issue for remote locations. So, whilst the process technology has generated the most interest, and rightly so when considering the need for optimum catalyst and reactor design, we must not forget that there are equal challenges in the support systems when considering engineering, construction and cost. The way to look at a GTL facility can be summarised as: l A world-scale gas processing and syngas generation facility at the front, together with at least two of the largest single-train air-separation plants ever built l Large-scale chemical conversion process in the middle l A refinery on the "back end" l

OFFSHORE WORLD OCT-DEC 2003 n 56

The Steam Systems

A GTL facility of around 34,000 bpd has a steam system of considerable size, which, depending on the technology selected for each of the processes, will have to handle a total steam rate of around 1,500 tonnes per hour. This level of steam handling will, no doubt, surprise the refiners, but will also surprise those used to dealing with syngas complexes, as this is approximately double what you would expect from the syngas capacity. For GTL, this steam will normally be of two discrete pressure levels – one associated with the syngas generation and one associated with the Fischer-Tropsch synthesis. The steam generated by the syngas processes is available at a variety of conditions, and is largely limited by what

ENGINEERING you feel able to generate with the syngas waste-heat boiler (levels are normally limited by metal-dusting concerns). However, the steam generated by the FT process is limited by the FT reaction conditions and is normally limited to less than 200 0C for liquid-phase processes. This steam should be viewed as an opportunity, as it gives unprecedented flexibility in configuring the utility systems: l High-pressure (HP) steam pressure and degree of superheat can be optimised for a particular cycle and steam turbine set. l Medium-pressure (MP) steam can be superheated, and due to the pressure level, this could be done easily using waste heat from another fired heater. l Use the HP steam for compressor drives, preheat or power generation. Used within the air separation unit (ASU), it is possible to have the main and booster air compressors on the same shaft w ith a single large turbine. l Use the MP steam for preheats or for power generation. Consider this for smaller compressor drives, if economically viable. l To reduce equipment, simply view the MP steam as an FT cooling medium and condense it. The MP steam has been a problem area with FT plants in the past, but considerably larger turbines are now available for using this valuable utility cost-effectively, again shifting the basis for what a GTL plant can achieve. This, coupled with more recent advances in mechanical design of gearboxes and complex shaft arrangements, gives the engineer even more flexibility in using this grade heat.

The Fuel System

FT plants unfortunately do not convert the synthesis gas into 100 per cent C5+ hydrocarbons. The combined FT reactor effluent is a cocktail of hydrogen, nitrogen, CO, CO2, water, water-soluble oxygenated hydrocarbons, methane and C2+ olefinic and paraffinic hydrocarbons. This stream exits the FT reactor in the vapour phase and is usually condensed, at which point a hydrocarbon-rich and a water rich phase are removed. This leaves a vapour stream for which the engineer has several options: l Burn as fuel

Recycle into the process Remove "useful" molecules such as hydrogen l A combination of the above. Each of the different GTL processes has different methods of handling this stream, including some once-through processes that propose simply combusting the entire stream in a specially designed gas turbine. However, all of these processes have the following in common: l This stream represents the single largest quantity of high-grade heat within the process (ignoring , for now, small vents and light ends productions in other units). l The stream is of low heating value, due to the CO2 and nitrogen it contains (typically <350 Btu/SCF), which represents challenges in designing a stable fuel system and combustors. GTL facilities, whilst exothermic overall, do have needs for high-grade heat, as this can be exploited more cheaply in general than, say, the FT steam. Users include preheat requirements for chemical conversion (in syngas generation and hydrocracking), steam superheating , large rotating equipment within the ASU (the fuel could be used for gas turbines), and power generation. The challenge is to have a good database of costs and ensure that you are targeting the highest value energy at the right user. l l

Capital Cost vs Process Efficiency

All engineers are used to the perennial debate over cost and efficiency. This debate is totally valid for GTL facilities, but it is worth remembering that the conventional answers are no longer true, and you should investigate what fits the particular economics of your project. The reasons for this relate to the unique process and economic constraints that exist within GTL: l The process rejects ~40 per cent of the feed energy as waste heat. Due to the significant investment in air or watercooling facilities, efficiency is a key issue in capital cost reduction. l The plants are very equipmentintensive; so, unfortunately, opportunities are limited to reduce cost of a large "thing" (such as a single turbine), but savings in utility cycles chip away at OFFSHORE WORLD OCT-DEC 2003 n 57

SOLUTIONS

all equipment items in the system and so reduce cost. l The plants are capital intensive, so a large portion (>50 per cent) of the cash flow is capital repayment. This results in the usual capital versus operating expenditure balance being tipped towards capital costs. The requirement to check life-cycle costs is, therefore, more important than normal, but it is essential that any optimisation of this type is done against hard economic data and decision-making indicators (net present value, NPV, rather than payback), which reflect reality and not some project quirk.

Engineering Challenges

Engineering of GTL facilities has progressed beyond just process engineering that typifies most pre-feasibility and feasibility studies. During these early stages of project development, engineering development outside of the process development has focussed on the following main areas: l Plot plan and piping arrangements l Construction philosophy – stick-built/ modular/barge mounted l Foundation and civil design/site preparation l Heavy lift studies (the plant includes some significant reactor vessels) l Risk management l Standards and specifications l Local development l Environmental impact. The layout of GTL facilities was identified early as being cost-critical due to the process intensity of the projects. This manifested itself both in terms of savings from laying out the large quantities of equipment involved, and also in minimising piping runs of very large gas and steam systems. This saves not only basic bulk material costs, but also minimises pressure drops and ensures delivery of utilities at intended conditions. The other main area of interest in these early stages was associated with the large reactors involved, with detailed assessments made into the foundation and civil requirements, and also the construction methodology for these large vessels. Throughout these exercises, a common thread has been whilst FT-based GTL plants contain proven technology throughout the process, not all these processes have been

ENGINEERING

SOLUTIONS

used together before, let alone actually constructed together. For example, whilst at first sight a plant of this scale and complexity would logically be stick built, it must be remembered that often remote locations are being targeted, with minimal local infrastructure and resources, and also the technical complexity of the plants requires a degree of specialist supervision. The modularised plant also presents some challenges, through the size of the modules required versus the number of modules, through to targeting module yards with the required capability and transport of very large modules and the construction sequence. For this reason, stick-built is the construction method of choice for developed sites, and modular designs are used for remote locations. In addition, the complex nature of the plant, and the specialists required from the technology providers for commissioning and start-up, makes this a schedule-critical task in more ways than one. The utility-intensive nature of the plant also makes provision of start-up equipment cost-critical, as these will tend to be shutdown once normal operation is achieved. For remote locations, early start-

an order of magnitude more important than "conventional" enterprises, such as refinery upgrades. At the FEED stage, a detailed study of all alternatives must be carried out so that the FEED package is well defined and all activities are fully specified. If these projects are being considered for lump-sum turnkey bidding, the FEED package must be sufficiently defined to allow for this contracting strategy. Failure to define requirements adequately will lead to elongation of the bidding process, possible bid recycle, delays, and failure of the engineering , procurement and construction (EPC) contractors to provide bids within the narrow range required to close out the project financing, ultimately leading to higher bids. For these reasons, the level of activity, and the experience and expertise of the FEED contractor must be considered early

20%

These are covered in more detail below. The sequence of construction work must be carefully reviewed to enable each different trade to commence work sequentially or with a short lead-time, again, to avoid site congestion. Numerous heavy lifts are involved for a GTL facility, some of which are absolutely critical to overall construction schedule. The plot must be so planned as to enable timely sequencing of these lifts, and minimise the heavy-lift window to reduce cost. However, this must not be overly constrained, so as to allow some flexibility in the construction sequence. The above again demonstrates the need for adequate definition during FEED for these issues to be properly investigated and resolved, with beneficial engineering carried out in key targeted areas, to ensure adequate EPC definition.

30%

Syngas production FT synthesis Product work-up

Figure 2: Indicative cost breakdown.

15% Other process units

up of certain areas of the plant, to act as service providers to the remaining sections, is also needed.

Utilities Offsites

Engineering & Construction

The above highlights some of the areas that have been studied at length in the course of GTL plant evolution. However, we are now in a situation where all the studies and feasibilities have been investigated, and now the designs must be confirmed. In the GTL arena, two large projects have recently completed basic (i.e., front-end engineering design, or FEED) design, with one now undergoing detailed engineering. For projects of these sizes to be controllable in cost and schedule terms, all of the issues studied must now be resolved. A project of this magnitude requires more than 20,000 engineering deliverables. To optimise the considerable efforts involved, correct sequencing of the engineering effort is essential. The timely exchange of vendor information is paramount to avoid recycling of technical data, which, for GTL projects is

15%

10% 10%

within the commercialisation cycle of the GTL technology by the technology provider, or all the hard work performed in the laboratory and with pilot plants will not translate into a bankable project entity. A good example of this is the plot plan. A well-thought-out and properly developed plot layout is a major contributor to the success of the engineering work. Whilst optimisation of the plot area, as highlighted earlier, is essential for cost, one should also study the constructability of the plant, to avoid "boxed in" construction difficulties, and also allow for construction work to progress on multiple fronts. The plot plan also defines the commissioning requirements, which, for GTL, are somewhat different than normal. OFFSHORE WORLD OCT-DEC 2003 n 58

Commissioning a GTL facility is an art all its own. Most projects are usually tagged as "construction-driven", with the emphasis on the activities and sequences required to bring the facility to mechanical completion. However, the unique nature of GTL projects, and the way that they have evolved to reduce cost, means that GTL projects should really be tagged as " commissioning-driven". No new facility can create the cash flows that the project financiers are looking for until such facilities are commissioned, started up, and producing the product as specified - plants that are mechanically complete do not generate products. GTL projects have partly reduced their capital costs by successfully integrating the useful

ENGINEERING

Fischer-Tropsch Conversion

Oxygen Natural gas

SOLUTIONS

Sasol gas conversion process.

Hydrocarbon product

Natural gas reforming

Wax product

Product upgrading

Naphtha Kerosene Diesel

Synthesis gas

energy produced by the plant back into itself. The downside is that start-up equipment has been cut back to the bare minimum. This presents some interesting challenges for the commissioning teams. In addition, the ASU, which can be considered as being at the heart of the utility network for these facilities, requires early startup to enable commissioning of the other part of the facility as it is a key utility provider. For early ASU start-up, certain elements of the steam system will, therefore, require early completion, commissioning and start-up. This dictates the construction schedule as well as the original plot layout. The steam systems are also key to commissioning activities that are being undertaken elsewhere within the facility. For the more remote sites with no outside battery limit services, the GTL facility must also provide its own power from the start, creating another commissioning challenge. The implications of having some of the key utility systems "live" whilst construction is being completed in other areas of the facility is significant. Approximately, one-quarter of the GTL work involves the ASU and its associated facilities. Therefore, the ASU design and construction work must be carefully integrated with the remainder of the plant. The core technology of slurry-bed-based FT GTL technology is the reactor and its associated equipment. These very complex pieces require very well defined specifications, drawings and layouts at the FEED stage. The EPC contractor cannot be expected to be fully familiar with these details

of GTL technology, as no large-scale commercial-scale slurry bed-based GTL facility has yet been constructed. It is likely that the third or fourth GTL project will ease the burden on the EPC contractor, but until then, it is up to the technology provider and FEED contractor to ensure that the years of experience are well-presented and developed sufficiently for the EPC contractor. This element is also again essential to ensure that sufficiently high-quality bids are obtained from the EPC contractors to enable closure of the project financing. The parameters one has to consider to ensure cost-effectiveness during construction are selection of the plot layout, availability of lay-down area, resource (skill/trade) availability, camp facilities and other logistical issues. All of these must be considered at the outset if cost and schedule implications are to be avoided. This is always good advice for any project and not specific to GTL, but again the technological complexity of GTL facilities requires a special kind of diligence to ensure that inappropriate "industry normal practice" is not blindly applied. Interfaces are also key to the success of a GTL project - both internally for the various technologies that are being integrated, and externally with the site and environment in general. In addition to these considerations, a GTL facility is usually sandwiched between an upstream project of some type (perhaps wellhead facilities, pipeline and a gas plant) and a downstream development (export facilities and OFFSHORE WORLD OCT-DEC 2003 n 61

perhaps some form of infrastructure and utilities development). Synchronising these interfaces represents considerable challenges to the project planner, as different contractors and even different owner teams will be involved. If the GTL facility is integrated into these enterprises in the form of process technology and utilities, the challenges faced are even greater. As a final thought on engineering, it is worth considering some of the quantities involved for a generic 34,000 bpd grassroots facility: l Equipment count ~400 (excluding vendor packages) l Average pipe diameter - ~8 inches As evidenced in Figure 3, the quantities involved are significant, making a grassroots ~34,000 bpd GTL facility roughly equivalent to a grassroots ~100,000 bpd refinery in engineering and construction terms.

Schedule

The optimum schedule for building a 34,000 bpd GTL facility is around 30-33 months from start of detailed engineering to mechanical completion. A further five months is required to carry the plant through pre-commissioning, commissioning, startup and performance tests up to commercial completion. This schedule can be shortened by carrying out some beneficial engineering prior to the effective date of the EPC contract, thus ensuring that selection of

ENGINEERING

SOLUTIONS

long-lead items is made prior to the contract award. A small premium may be attached to this method of execution, but an overall analysis of the cash flows shows that this is money well spent on an NPV (net present value) basis when considering overall schedule implications. Other methods that can be used to shorten the schedule are: l Pre-assembly of units, either offsite and then delivered, or onsite prior to installation in the required area l Modularisation l Vendor alliances for unusual or specialist equipment items - in some area, GTL facilities require equipment items that are outside the normal range of items that vendors supply, or require

projects will develop in technology and engineering terms. The GTL projects to be executed towards the end of this decade are likely to have some startling differences to the current ones as technology advances are made, and as we build upon the project execution and operational experiences that will result. Next-generation GTL facilities will probably differ in two distinct areas, over and above the general "creeping" improvements in machinery and catalysts. The first, and most likely area, which we have already seen to some extent, is in economies of scale. The current crop of GTL facilities are considering multiple trains, with the slurry bed technologies in the ~15,000 bpd train size. We are likely to see

The second and exciting area of development is in the syngas generation step. Developments in novel approaches, such as ceramic membranes, would be a stepchange in technology terms, allowing syngas generation without an air-separation unit. In the nearer term, however, we are likely to see projects using gas-heating reforming. Several types of this technology are available, with several others in development, but all sharing the common characteristic allowing the heat in the synthesis gas to be "recycled" back into the pre-heat or conversion steps. This removes or reduces the need for large-scale highgrade energy use in preheating , and simplifies waste heat recovery from the syngas in generating steam. These technology and equipment developments, together with continued advances and experience in engineering Figure 3: Bulk material quantities.

600,000

500,000

400,000

Concrete (m3) Paving & gravel (m2)

300,000

Insulation (m2) Piping (m)

these amazing projects, will ensure that GTL is indeed the launch pad to a new hydrocarbon future. The EPC contract for Qatar Petroleum and Sasol's Oryx GTL project at Ras Laffan, in Qatar, has now been awarded, making this technology the most significant advance in gas processing of the new millennium.

Welding (dia ins)

200,000

Cabling (m) 100,000

0

specialist engineering efforts, due to unusual application (low-pressure saturated steam or lower heating value fuel are two common examples).

The Future For GTL

If we take the current crop of GTL projects as having been successful in their engineering and economic development, and these are considered to be robust enough to attract project financing, we can state that no current barriers exist to current GTL project implementation. These achievements must be applauded, but we must consider how the

dramatic increases in size to beyond 20,000 bpd, with developments in: l Slurry reactor design, fabrication and erection l ASU sizes, "springboarding" off increased air compressor sizes l Greater confidence in terms of reliability with single-train utility systems, with potentially very large de-aerator and condensate-handling facilities l Large steam turbines, able to deal with FT-derived steam l Gas turbine developments, both in size and design, for low British thermal unit (Btu) gases. OFFSHORE WORLD OCT-DEC 2003 n 62

Mr Simon Clarke is Manager (Gasto-Market Technology) for Foster Wheeler Energy Limited, based in Reading, UK. Since 1997, his main activities have been in the GTL arena, playing a leading role in Foster Wheeler's team supporting Sasol in their GTL developments and the recent front-end engineering activities for the Qatar and Nigeria GTL projects. Contact details: Foster Wheeler Energy Limited, Shinfield Park, Reading, Berkshire, RG2 9FW, UK; E-mail: [email protected] Mr Bahram Ghaemmaghami is a Project Director with Foster Wheeler Energy Limited, also based in Reading, UK. He is responsible for Foster Wheeler's current GTL projects, including Qatar (Ras Laffan) and Nigeria (Escravos). His contact address is same as that of Mr Clarke. E-mail: [email protected]