AUSTRALIAN LIQUIDS-HANDLING SYSTEM CUTS SURGES TO LPG PLANT

Aug. 6, 1990
Geoff McKee Terry D. Stenner Bridge Oil Ltd. Sydney A pipeline liquids-handling facility recently commissioned by Bridge Oil Ltd., Sydney, allows gas production to be quickly ramped up to meet customer demand. Its design eliminates troublesome liquid surges which had hampered plant operations.
Geoff McKee
Terry D. Stenner

Bridge Oil Ltd.
Sydney

A pipeline liquids-handling facility recently commissioned by Bridge Oil Ltd., Sydney, allows gas production to be quickly ramped up to meet customer demand. Its design eliminates troublesome liquid surges which had hampered plant operations.

The pipeline-loop system, located at the Wallumbilla LPG processing plant, Queensland (Fig. 1), was built for 60% of the cost of an equivalently sized conventional slug catcher. Its control system enables automatic, unattended handling of liquid surges and pigging slugs from the 102-km Silver Springs to Wallumbilla two-phase pipeline.

Because of this system's simple hydraulics, normal slug-catcher piping design problems are eliminated. Safety is improved because the potentially hazardous condensate liquid is contained in a buried pipeline.

PIG RUNS

Slug catchers are vessels which provide high-pressure liquid surge capacity at the downstream end of a two-phase gas pipeline. An increase in gas flowrate may cause such a surge by sweeping out additional liquid.

A pig or sphere run through the line creates the largest slug and usually determines the size and cost of the slug catcher. Regular pigging lowers pipeline pressure-drop which increases the useful capacity of the pipeline.

Free water and corrosion products are also removed by pigging, which keeps the pipeline clean and allows the distribution of corrosion inhibitor. This reduces the possibility of internal corrosion or blockage by hydrate formation.

On behalf of a number of joint ventures, Bridge Oil operates oil and gas production facilities in the Surat-Bowen basin area of Australia (Fig. 1). The LPG-processing plant, owned by Bridge Oil, Petroz, and AGL Petroleum, is located near the small town of Wallumbilla, 460 km (285 miles) inland from Brisbane.

Raw gas and associated condensate are fed to the plant through a 219 mm (8 in.) diameter pipeline from Silver Springs, 102 km (63 miles) to the south. At Silver Springs, Bridge Oil operates gas gathering, vapor recovery, and dehydration facilities.

After extraction of LPG in a turboexpander and fractionation process (Fig. 2), the residue gas is recompressed and sent through the Roma-Brisbane and Wallumbilla-Gladstone pipelines to industrial and domestic customers.

The LPG plant, operating at about 65% capacity, produces approximately 22 MMscfd of sales gas, 120 tons/day of LPG, and 48 cu m/day (300 b/d) of C5 + condensate.

Before construction of the LPG plant in 1984, only minor quantities of liquid formed as a result of retrograde condensation in the Silver Springs-Wallumbilla pipeline. When the LPG plant was built, vapor-recovery facilities were installed at Silver Springs.

LPG-rich separator vapors, previously flared, were recovered and compressed into the pipeline causing it to operate in the two-phase flow mode. Regular pigging had to be temporarily suspended due to a lack of slug-catching facilities at Wallumbilla.

Available correlations which predicted liquid holdup gave widely varying results (Fig. 3).2

Based on these curves, the size of a liquid surge resulting from a rate change varies between predictions by a factor of five. The same curves show a disagreement, by a factor of three, in the volume of liquid arriving at the plant ahead of a pig.

A slug catcher, designed on this basis, has a high probability of being incorrectly sized. Operating problems caused by an undersized slug catcher would be overcome by pigging at short intervals to reduce slug size. This would itself be inconvenient and reduce manpower efficiency.

An oversized slug catcher would have reduced the favorable LPG plant economics.

Consequently, in early 1985 the slug-catcher project was deferred until actual pipeline data and plant operating experience enabled more accurate definition of the pipeline's requirements.

OPERATING PROBLEMS

Free-liquid slugging into the LPG plant dominated operating concerns from start-up in 1984 until commissioning of the new facilities at the end of 1989.

During this period, the plant's inability quickly to ramp up sales-gas production necessitated assistance from other producers who share the market. They would be asked to supply the temporary shortfall until the slug problem eased.

De-ethanizer flooding often occurred when slugs had to be processed to meet customers' gas requirements. This severely affected distillation efficiency, causing propane product to be "lost" in the overhead sales gas.

When very large liquid slugs arrived, liquid had to be flared to avoid a plant shutdown.

Corrosion products from the pipeline, containing iron sulfide and mixed with glycol carry-over from the field dehydrator, arrived from time to time with incoming liquid. On several occasions, this mixture contaminated the liquid molecular-sieve bed, allowing moisture to enter the cryogenic section of the plant.

A costly exercise of defrosting followed, resulting in downtime and lost production.

The inferred pipeline internal corrosion could only be corrected by resuming pre-1984 regular maintenance pigging. A lack of slug-catcher facilities also prevented the running of an intelligent pig through the pipeline to assess the degree of corrosion, if detectable.

In September 1988, gas production was forecast to increase. Bridge Oil management directed that a detailed slug-catcher cost estimate and project justification be prepared.

As a first design step, a brief survey of local slug-catcher installations was carried out.

The largest two-phase, raw-gas pipelines operating in Australia are Woodside Petroleum's 1,016 mm (40 in.) diameter, 135-km (83-mile) North West Shelf offshore pipeline; Esso/BHP's 508 mm (20 in.), 108-km (67-mile) Marlin offshore pipeline; and a 762 mm (30 in.) diameter, 104-km (64-mile) trunk, line operated by Santos in the Cooper basin.

Multiple-barrel slug catchers with a liquid holding capacity of 5,200, 840, and 600 cu m, respectively (32,500, 5,300, and 3,800 bbl) were installed to serve these pipelines.2 3

A single-barrel slug catcher is suited to smaller two-phase pipelines, where less than about 150-cu m (1,000-bbl) liquid-storage capacity is required. Sizes range from 1,016 to 1,219 mm (40-48 in.) diameter and up to 150 m (500 ft) long.

Similar units have been installed in the Cooper basin's Tirrawarra production facilities operated by Santos and, more recently, at the Kincora LPG processing plant operated by Oil Company of Australia.

DRAWBACKS

Although made from large-diameter linepipe, aboveground slug catchers are classified in Australia not as "station pipework" under the Pipeline Code (AS2885) but as pressure vessels. Accordingly, their design, fabrication, testing, and inspection must comply with the Unfired Pressure Vessel Code (AS1210).

The stringent requirements of the pressure vessel code, combined with the need to import large-diameter linepipe into Australia, amounts to a significant incremental cost.

Alternatively, large tailor-made pressure vessels fabricated from boiler plate have been used4 but cost considerations would normally bar this option.

Because slug catchers contain large amounts of highly flammable liquid at high pressure-normally around 7 MPa (1,000 psi), extensive measures must be taken to reduce the risks and consequences of fire.

The primary means of fire protection involve prevention of gas leaks and ignition sources.

Secondary means are pressure-relief valves, automatic gas and fire detection, fixed water-spray systems or firewater monitors, emergency shutdown valves, fireproofing of vessel and pipe supports, and good drainage under the vessel. Fencing, lighting, and access roads are also important safety considerations.

In Australia, both vessel and fire-protection design must be approved by the designated state government authority. The required safety systems add further to an aboveground slug catcher's initial capital and ongoing maintenance costs.

All these safety systems add further to the initial capital and ongoing maintenance costs.

The large space required for a conventional slug catcher was another disadvantage for our project because the available plot area was severely restricted.

Complex hydrodynamic studies are sometimes required to protect against design defects that can occur in conventional slug-catchers. Such defects are uneven liquid distribution, liquid carryover, and excessive forces on bends and tees affected by the liquid slug.

For some projects, even performance of laboratory models is studied before commitment to a particular design.2

For these reasons, there was general dissatisfaction with the cost of a conventional slug catcher, expressed as dollars per unit volume of capacity, especially when compared to underground liquid pipelines.

AN ALTERNATIVE

Several weeks were spent investigating ways to reduce costs and improve inherent safety. Both earth-mounded and buried single-barrel slug catcher options were assessed to find out whether reduced pressure-relief and fire-protection requirements had any significant impact on overall cost.

A single aboveground separator, linked to a series of underground liquid-storage bottles fabricated from surplus 24-in. API linepipe, was estimated.

The use of a loop line satisfied all our requirements. The concept grew from a description of the mode of operation employed by Esso Australia to pig the Marlin pipeline.3

For debottlenecking reasons, this pipeline had been looped for approximately 20 km upstream of the Gippsland gas-processing plant near the town of Sale, Victoria.

For the Marlin line, it was reported that the amount of condensate brought in by a pig would always exceed the capacity of the slug catchers. Therefore, how much condensate was taken out of the pipeline must have been limited to avoid overfilling the slug catchers and exceeding the processing capacity of the plant.

Designing the slug catchers large enough to hold all the condensate removed from the line when the pig was sent through was impractical. Most plants could slow the pig by reducing their inlet-gas rates.

In the case of the Gippsland plant, however, which was the sole gas supplier to a major gas utility, this option is unavailable. Even during a pig run, gas rates had to be maintained to satisfy the demand.

This aim was achieved by opening the cross-over valve at Site 3 after the pig had passed to allow gas from the line being pigged to proceed up the other line.

Condensate in the line being pigged was fed to the plant until the line was cleared.3

It was apparent that the last section of Bridge Oil's pipeline could be looped in the same way.

The new loop could be sufficiently long to enable the adjacent parallel section of the existing line to store a full-sized pigging slug, which would be processed at leisure while gas was being supplied to the LPG plant through the loop line.

A new pressure vessel would be required to receive the slug but only of sufficient size for gas-liquid separation and flow control. Without the loop line, gas supply to the LPG plant would be interrupted for up to a day while the slug was processed.

Another way of viewing the function of the loop line is to see it as a back-up source of gas during pigging interruptions, not from an obliging neighbor, but from one's own pipeline.

The pipeline loop liquids-handling concept is apparently not new. According to one report: "Another technique offshore involved a mile or so of parallel underwater liquid pipeline and switch-gear that could be likened to a sidetrack on a railroad, deadheading the liquid so the gas could get by."5

The challenge was to design an automatic control scheme. Piping design was comparatively straightforward (Figs. 4 and 5).

FUNCTIONAL SCRAP

A surplus refinery pressure vessel, to be sold for the price of scrap metal, was brought to our attention (Fig. 6a). It was 1,500 mm in diameter and 5,500 mm long (5 x 18 ft) and built to ASME code with a design pressure of 14.5 MPa (2,100 psi).

After a favorable internal inspection report, the vessel was purchased, refurbished, and recertified for use on this project (Fig. 6b). This saved 80% of the cost of an equivalent tailor-made vessel.

A pigging test was carried out on the pipeline in October 1988. Temporary separators and atmospheric condensate tanks had to be set up at the LPG plant to catch resulting liquids.

The test pig swept 93 cu m (585 bbl) of condensate together with 4 cu m (25 bbl) of water-glycol from the pipeline. Gas rates before the test were 622,000 standard cu m/day (scmd; 22 MMscfd), showing that the correlation of Eaton (Fig. 3) most closely predicted liquid holdup.6

A high point on the existing pipeline, 3.7 km upstream of the LPG plant, was selected as the loop line tie-in point (Fig. 4). This gave liquid-storage capacity of 125 cu m (790 bbl) in the parallel main line. A cost estimate was carried out based on a preliminary piping and instrument diagram, plus plot plan. Results indicated approximately $500,000 could be saved using the loop-line scheme in preference to a conventional slug catcher of the same capacity.

The savings would be approximately $120,000 less without acquisition of the refurbished pressure vessel.

Two aspects needed to be addressed in the project's "Authorization for Expenditure" submission to joint-venture partners:

  1. The need for the project itself-The justification was based on qualitative judgments concerning pipeline maintenance, security of gas supply, operational safety, pipeline and plant capacity, plant operability, and relations with Surat coproducers.

    It is difficult, if not impossible, to calculate clear-cut economic yardsticks for a project of this type.

  2. The unusual design-Without any direct knowledge or experience of similar systems, the loop line justified itself on grounds of reduced capital cost. The cost savings were considered to offset any uncertainties resulting from a design which had not been tried and proven.

P&ID DEVELOPMENT

Following project approval in February 1989, the main design effort was directed to the piping and instrument diagram (P&ID).

This diagram brings together all crucial design ideas; defines the control scheme; determines safety, operability, valve, and instrument requirements; and ultimately governs the final cost of the project.

A primary concern was how to ensure success of the control scheme. A joint-venture approach between engineering and operations staff was adopted in order to incorporate the control requirements of those most affected by the decisions made. This also would ensure common responsibility for, and acceptance of, the end result.

Fig. 7 depicts, in simplified form, the control scheme which was finally adopted. The key features are a gas-operated ball valve (FV) and electronic pig sensor (XS) at the loop line tie-in and two _P-cell level transmitters (LTs) on the liquid receiver-separator vessel.

The remote valve is operated directly by the pig signal or from the plant's control room by telemetry. The statuses of both the valve position and pig sensor are displayed on the control-room telemetry panel. Power supply to the remote telemetry station is provided by a 12-v solar panel battery unit.

Both level transmitters span 0-100% of the vessel's liquid-level range. The two level switches on the vessel ensure proper calibration of the _P-cell transmitters.

As shown in Fig. 8, one level transmitter is devoted to controlling level by ramping the setpoint of the liquid outlet flow controller (FRC).

The other level transmitter takes over when this control loop cannot maintain a level due to excessive liquid inflow. Its output goes to a control block (LIC) which maintains a preset level in the vessel by adjusting output to the split-range gas flow control valves (FCV-1 and FCV-2).

These valves automatically adjust the balance of gas flow between the main line and loop line. When a pig is being brought in, the vessel gas-outlet valve (FCV-2) tends to close with most of the gas being supplied to the plant through the loop line. The flow transmitter on the vessel's gas-outlet line is for indication only.

Detailed design, procurement, and construction were carried out during 1989. The project was commissioned on schedule in December 1989.

FACILITIES' OPERATION

Since commissioning, the facilities have eliminated all the operating problems associated with liquid slugging into the plant.

Small day-to-day liquid surges are automatically flow controlled by the lower level control loop previously described.

Gas flow through the loop line can be used to maintain a steady flow of liquid into the plant if larger slugs or continuous high liquid rates are being experienced. The loop line tie-in rises 2 m vertically and is located on a high point.

Very good gas-liquid separation occurs at the loop line tee, and liquid remains in the main line. This is believed to result in part from the horizontal momentum of liquid droplets in the main line.

Under these circumstances the control-room operator has two choices to control liquid entering the plant.

He may manually open FCV-1 and slow down liquid arrival at the vessel by supplying some plant gas requirements through the loop line. The lower velocity in the main line increases liquid holdup in that line allowing it to serve as a surge vessel.

Alternatively, he can hold the vessel's level steady by running the upper level control loop (LIC) on auto at a level between the LAH and LAL ramping points. The vessel's liquid outlet flow is manually set to the desired rate on the FRC, and the LIC will control the loop-line by-pass gas rate to hold a steady vessel liquid level.

In the event of the vessel's level falling with no gas flow through the loop line, then the low level switch (LAL) will ramp down liquid flow to the plant as necessary. Any pipeline corrosion sludge which enters the vessel is transferred via a drain line to a field mud tank for safe storage before disposal at a later date.

Preparation for the arrival of the pig at the remote valve station involves resetting the pig signal and closing the remote valve (FV) to avoid the possibility of liquid entering the loop line. As the pig passes the valve station, the pig sensor automatically opens the valve. The statuses of both the pig sensor and the remote valve are transmitted by telemetry to the control room.

Processing the liquid slug ahead of the pig is handled automatically in the following manner:

The vessel level controller (LIC) is set to control level at 50%. Increasing output opens FCV-1 and closes FCV-2. When output from the level controller is 50%, both FCV-1 and FCV-2 are fully open.

The LIC is backed up by a high level trip (LAHH) which, at 80% level in the vessel, sets the output from the LIC to 100%. This fully closes the gas outlet from the vessel (FCV-2) and fully opens the loop line (FCV-1).

An interlock holds the LIC and LAHH output at zero until a telemetry signal is received, indicating that a pig has passed the pig sensor and the remote valve (FV) has opened.

The control-room operator sets liquid flow into the plant on the flow controller (FIC) at a rate suitable for smooth plant operation. The level controller (LIC) splits incoming pipeline flow between the loop line and pigged line to keep liquid arriving at the preset rate.

Under these conditions, the pig arrives in the plant trap 24-48 hr after passing the loop-line valve station.

The arrival of the pig in the plant trap is indicated by a fall in the vessel's liquid level and the level controller (LIC) closing off the loop line completely. Any corrosion sludge which sometimes accompanies the pig's arrival must be manually drained to a mud tank from the vessel's bottom drain.

Operational flexibility was recently demonstrated when the plant was unable to deliver LPG due to lack of customer storage capacity. Several hours after a pig was launched at Silver Springs into the raw-gas pipeline, the LPG purchaser advised LPG takes were curtailed.

The plant staff merely opened the loop line 100% after the pig had passed by the tie-in point. A large slug was held in the last section of main line for several days until LPG takes resumed.

Because the gas behind the pig is liquid-free for several days, LPG production was delayed by the pigging operation.

WHAT'S BEEN LEARNED

This liquids handling design has been tried and proven for our particular line size and capacity. It could be scaled up and used on bigger projects, resulting in similar operational advantages.

Depending on local cost factors, considerable capital cost savings are possible.

If the loop line and main line are constructed simultaneously, as for a new development, costs will be further reduced.

Due to this system's simple hydraulics, normal slug-catcher piping design problems are eliminated. Safety is improved because the potentially hazardous condensate liquid is contained in a buried pipeline.

ACKNOWLEDGMENT

The authors wish to thank the joint-venture participants, Bridge Oil Ltd., Petroz NL, and AGL Petroleum, for their permission to publish this article.

We also wish to thank colleagues in the Bridge production department for their assistance and comments, particularly A. A. Young, R. D. Frith, J. C. Dowell, and T. J. Petersen.

REFERENCES

  1. Frith, R. D., "Performance of the Wallumbilla LPG plant," APEA Journal, Australian Petroleum Exploration Association Ltd., 1986, Vol. 26, Part 1, p. 485.

  2. Seymour, E. V., Craze, D. J., and Ruinen, W., "Designing Australia's North West Shelf offshore pipeline," OGJ, May 7, 1984, p. 128.

  3. Cunliffe, R. S., "Condensate flow in wet-gas lines can be predicted," OGJ, Oct. 30, 1978, p. 100.

  4. Huntly, A. R., and Silvester, R. S., "Hydrodynamic analysis aids slug catcher design," OGJ, Sept. 19, 1983, p. 95.

  5. Martin, R. E., "Handling liquids in offshore gas lines gets new approach," OGJ, Apr. 27, 1981, p. 143.

  6. Eaton, B. A., Andrews, D. E., Knowles, C. R., Silberberg, I. H., and Brown, K. E., "The Prediction of Flow Patterns, Liquid Holdup and Pressure Losses Occurring During Continuous Two-Phase Flow in Horizontal Pipelines," Journal of Petroleum Technology, June 1967, p. 815.

Copyright 1990 Oil & Gas Journal. All Rights Reserved.