OFFSHORE TOPSIDES-1 Decision trees aid production facility design

John J. MacDonald
Chevron Petroleum Technology Co.
San Ramon, Calif.

Robert S. Smith
OPC Engineering Inc.
Houston

This methodology rapidly evaluates crude/gas production topsides for reducing overall facility installed cost.

Preinvesting in offshore topsides for handling upside uncertainties in reservoir withdrawal may significantly improve the internal rate of return (IRR) for the overall field development.

An offshore field development plan for an incremental or new project ideally consists of the following:

• Optimize reservoir engineering

• Optimize petroleum engineering completion strategy

• Minimize flow line/pipeline or other transportation costs

• Minimize subsea or support structure costs

• Maximize present worth by ensuring the highest revenue stream

• Evaluate the risk/present worth relationship

• Minimize manning relative to overall present worth and risk

• Minimize overall operating costs

• Consider company practices/culture.

This first in a series of two articles focuses on production and test separation systems installed on a platform.

Topsides design

Most field development planning models rely only on cost spreadsheet models, which cannot demonstrate the cascading sensitivity of an integrated process, simulation/weight estimation model. The elements of design can be related to a decision tree methodology to reflect such sensitivity.

Until about 1985, value judgments and historical data formed the basis for parametric studies. During this period, topside facilities were designed by specific guidelines relative to various company cultures. Value judgments determined the number of spare units and production train configurations.

The price shock of 1986, when crude price decreased in a few months to $10 from $28/bbl, changed most of these leisurely, value judgment attitudes.

Since 1985, economic development of marginal fields has required facility engineers to discard past company culture bias for equipment configuration. This attitude adjustment is now embracing technology options that were considered too risky only a few years ago.

One such method involves using such models as OPC*MAP to provide additional quantitative guidance to qualitative decision trees. This model combines an integrated process simulation, weight estimation with capital and operating cost determination.

The main production facilities and reservoir fluid treatment can encompass half of an offshore topsides platform area and be half of the capital/installed cost. Manning levels may also, in some cases, represent substantial size and costs that can be reduced with an incremental present worth model.

Separation trains

The OPC*MAP software program compared various alternative separation schemes. This comparison is for Gulf of Mexico costs but other regions in the world will still have the same overall trend, although with different absolute costs.

Fig. 1 [21184 bytes] shows a generic multistage separation flow scheme. In Fig. 2 [18287 bytes], costs are compared for installing a three-stage separation facility having 1, 2, and 3 trains with production per train of 120, 60, and 40%, respectively. The analysis considers all of the cascading weight, area, and utility impacts.

A 2,000-scf/bbl GOR and separator stage pressures of 500, 200, and 15 psig were assumed. All gas was assumed compressed for export at 1,200 psig.

Fig. 3a [81923 bytes] relates a decision tree for the number of production trains based on value judgments from experience. The results can be confirmed with several case histories and parametric studies using OPC*MAP, which generates equipment lists of sized equipment.

The following factors must be considered when selecting the number of separation trains:

• Flowing well head pressure relative to the number of producing horizons

• Separator throughput

• Selection of vertical or horizontal separators

• Gas/oil ratio

• Wax content

• Sand content

• Turndown ability

• Separator availability

• Module clearance heights

• Ease of transport of large diameter separators, greater than 14 ft.

Production horizons

The number of required separation trains is greatly influenced by the number of producing reservoir horizons. In fields producing more than one horizon, if the flowing well head pressure (FWHP) of one horizon cannot match the FWHP of another horizon, the field may need a second separation train.

But if the available FWHP from the second horizon matches the second-stage separator pressure, it may be possible to handle production from both horizons in a single separation train.

Throughput

Oil, gas, water throughput volumes also influence the number of separation trains. Crude will usually have a plateau period, and water production and gas/oil ratio will change over time.

In general, a single separation train will be sufficient for fluid rates of 150,000 b/d. This is based on a three-phase horizontal vessel, 12-ft diameter by 45-ft long, with a fluid GOR of 600 scf/bbl and a 28-35° API gravity oil.

Physical size and lifting weight are two limits for installing production separators.

Separator type

The area, weight, and increased instrumentation for two or more separation trains is difficult to justify for offshore facilities. Two separation trains should be avoided if the horizontal separators cannot be piggybacked. In general, there is little economic advantage for having two separation trains unless they can be piggybacked horizontal separators.

If the process permits a vertical separator, throughput volumes for a single train are somewhat less than for a horizontal separator and will require a detailed study based on actual reservoir fluid composition.

Very often, vertical separators are installed when the reservoir fluid has a high GOR and the field produces with relatively few wells.

Two vertical separation trains may be required if the field produces from several horizons.

GOR

Gas/oil ratio is one criteria for determining vertical or horizontal separator diameters. GOR will also influence the decision on whether to retain a single separation train.

At higher GORs, vessels require larger diameters for the same amount of crude produced. This can be quickly shown by the "gas controlling" and "liquid residence time controlling" vessel sizing equations.

Wax content

Wax content may influence the number of separation trains. Production could be interrupted by having to shut down a separation train if wax build-up occurred and the vessels needed to be steamed or cleaned.

Thus, if wax content is high and processing conditions require external heating to reduce upsets, more than one separation train might be considered, especially in cases where the external heating might be interrupted.

Sand content

Regular separator cleanout may be required if the reservoir fluid sand content is severe and not controllable by gravel packing at the reservoir face.

Under these conditions, more than one separation train would avoid interrupting crude production.

Turndown

A large single-train separator may have a turndown (ability to operate at low flow rates) problem. The problem can be overcome by installing dual control valve sets on the liquid and gas outlets, sized to accommodate the entire flow rate range.

To accommodate low flow rates, a test separator can serve as a start-up separator until the crude rate merits a large single-train separator.

Availability

More than one separation train may be justified if the reservoir production potential is uncertain and an overdesigned topside facility has minor overall economic impact. This is a value judgment and beyond the scope of most engineering calculations.

The economic impact of two or three trains can be quickly evaluated with such software as OPC*MAP.

Individual components can be evaluated as mathematical probabilities. The results can point to having more than one separation train available.

This type of evaluation has not, in the past, been an overwhelming reason for having two or more trains. Other considerations, discussed in this article, will affect the number of trains on a less subjective, more quantitative basis.

Modules clearance

Individual separator stages may be "stacked" in a piggyback manner, particularly if separators are horizontal. Module clearance height is a restraint for this option.

Module clearance may not be a significant factor except for fields producing at high flow rates. The facilities should not be inhibited by structural restraints. If facilities are designed in parallel with the topside structure, the limits can be adjusted by the structural engineer.

An exception to this is the structural restraint imposed by a jack up drilling rig. The taller the deck height clearance, the greater the topside weight.

Two 12-ft diameter vessels may be stacked with a total module clearance height of 32 ft.

For GORs greater than 300-500 scf/bbl, more than two separation stages may be needed to retain heavy end components and to minimize gas compression requirements.

Separation stages

The number of separation stages is primarily influenced by export conditions and reservoir fluid properties.

The reason for more than one separation stage is to retain as many heavy ends, C5, C6, C7+, in the crude as possible by flashing reservoir fluid in a stepwise pressure reduction.

If produced gas is being recompressed for treatment prior to export or reinjection, compression requirements may be minimized by having high-pressure separation stages.

Crude export

Crude is usually exported by tanker as a degassed liquid, or by pipeline as a "pressured crude," to be further degassed onshore. The impact of the crude export specification on the number of separation stages depends on the gas present, crude temperature, and GOR.

For a tanker with a normal specification of 10-12 Rvp crude, the last-stage separator temperature would be about 130° F., with an operating pressure of about 20 psia. A minimum of two separation stages is normally required for this tanker condition.

The export Rvp may be related to the pressure (psia) and temperature (°F.) of the last separator stage by the following equation:

Rvp = 0.0532 X P X exp(532.5[1/(T+459.6) - 1/559.6] - 0.436)

Reservoir fluid properties

Three reservoir fluid properties impacting the optimum number of separation stages are: GOR, water content, and available FWHP. The influence over time of increasing GOR, increasing water content, and decreasing FWHP are also considerations for selecting the number of separation stages.

Table 1 [8929 bytes] relates the number of separation stages to GOR and Rvp.

Higher flowing well head pressures are required to support multiple stages. These higher pressures are usually available with higher GOR crudes. Separation pressures may be maintained by formation water drive or gas cap support.

If the separated gas is compressed, one way to minimize the number of stages required for recovering heavy ends is to recycle condensate from the compressor intercoolers back to the production separators. The number of separation stages is then a straight-forward economic comparison of horsepower and fuel gas saved in gas compression compared to the expense of additional separation stages.

Again, with software such as OPC*MAP, case studies can be performed rapidly on specific reservoir fluids to evaluate overall investment costs.

The crude's water content may affect the number of separation stages. Very often crudes with low GORs have a high water content. In this case, a single "bulk" separator often is installed. The remaining oil is then treated in one or two trains for water removal. A sufficient residence time is needed to remove enough water for attaining the crude export quality.

Potential well head pressure decline greatly impacts the number of separation stages. On new fields with unknown reservoir decline properties, conservative value judgments are often made for selecting the number and pressure of separation stages.

For example, bottom hole reservoir fluid samples from one North Sea development indicated a GOR of 600 scf/bbl and a 1,200 psi FWHP at the designed flow rate. The conservative facility design included two separation stages at 50 psig and 5 psig, but the field eventually produced a 1,400 scf/bbl GOR crude.

A high-pressure "bulk" separator with a minimum of residence time could have minimized compressor horsepower without much impact on overall investment. But the "conservative" decision on anticipating declining reservoir pressures and selecting a minimum number of separator stages at low pressure was costly to crude production potential.

A more economical gas compression configuration would have allowed more crude production without being limited by the gas rate.

Fig. 3b [81923 bytes] shows a decision tree for determining the number of separation stages. Fig. 4 [22242 bytes] illustrates OPC*MAP calculated installed costs for a different number of separation stages of conventional crude at two flow rates.

Test separators

Typically, a test separator is a single vessel used to determine, over time, the oil, water, and gas properties from each producing horizon or well.

If the producing field extends over the property lines of several co-owners, the test separator serves to determine the relative revenue payment to each co-owner.

A test separator may not be needed if two separation trains are available. Allocations for any single well may then be obtained by calculating differences under steady-state flow conditions.

Without two separation trains or if rate differences are not acceptable, the installation may require a "minimum" test separator. Multiphase flow meters are also a possible alternative.

Fig. 5 [34239 bytes] illustrates a decision tree for selecting a test separator.

Conventional test separator

Conventional test separators may be horizontal or vertical and sized for the maximum well potential and anticipated future gas and water rates. Test separator operating pressures are the same as in the first stage of the main separation train.

Conventional test separators usually measure three-phase flow (oil, gas, and water). Their size is normally fixed by the residence time required for oil/water separation. Because of the separator's cyclical nature, the shutdown sensors and actuators are at least equal to and perhaps exceed the range and accuracy of the main separation train controls.

Minimum test separator

A minimum test separator is only a two-phase vessel (liquid and gas), that separates and measures both phases.

A densitometer can measure the liquid density. The following equations can determine water content:

 Vw + Vo = Vm
 ÞmVm = ÞoVo + ÞwVw
 Vw = Vm(Þm - Þo)/(Þw - Þo)
 Vo = Vm - Vw
 where:
 Þo = Oil density, saturated with gas, calculated from lab tests
 Þw = Formation water density, known from lab tests, corrected for temperature
 Þm = Mixture density, measured by densitometer in the field
 Vm = Mixture mass flow rate, measured in field by high precision device
 Vw = Water mass flow rate
 Vo = Oil flow rate.
Multiphase flowmeter
 Multiphase flowmeters have been investigated and tested for several years and are candidates for replacing test separators in certain situations.
The Authors
John J. MacDonald is a senior facilities engineer with Chevron Petroleum Technology Co., San Ramon, Calif. He is involved in researching and applying technologies to reduce project life-cycle costs and improve production from offshore production facilities.
 MacDonald has a BS in civil engineering from the Nova Scotia Technical College and a BS in chemistry from St. Francis Xavier University.
Robert S. Smith founded OPC Engineering in 1979. He has worked on many oil and gas projects throughout the world. Smith has a BS in chemical engineering from the University of Tulsa. He holds four patents on gas and sulfur processing technology.
Copyright 1997 Oil & Gas Journal. All Rights Reserved.