Injecting acid gas with water creates new disposal option

ACID GAS DISPOSAL-1

Dave Kopperson, Steve Horne, George Kohn,
Dwayne Romansky, Chong Chan
PanCanadian Petroleum Ltd.
Calgary

Gordon L. Duckworth
DPH Engineering Inc.
Calgary

PanCanadian first tested acid-gas disposal with produced water injection at its David Battery No. 3 near Provost, Alta. (Fig. 2 [8,539 bytes]).
The second site for acid-gas disposal with produced water injection was at PanCanadian's Thompson Lake facility near Hardisty, Alta. (Fig. 3 [10,366 bytes]).

A scheme for injecting low-volume acid gases in an aqueous phase into a disposal well has decreased acid-gas disposal costs and facilitated operations in two Canadian fields.

PanCanadian Petroleum Ltd. and DPH Engineering Inc. developed the scheme.

With tightening environmental policies,¹ many companies are investigating alternatives to atmospheric sulfur and greenhouse gas emissions.

Alberta's oil and gas industry typically recovers a high percentage of sulfur in large, sour-gas processing plants, but has looked for more cost-effective approaches when dealing with small-volume plants.

This first in a series of two articles describes the acid gas/water injection process and laboratory testing. The concluding part will discuss field results.

Advantages

The PanCanadian and DPH Engineering scheme for disposing of acid-gas with oil field produced water offers the following advantages:

Eliminates high-pressure, acid-gas pipelines, dehydration systems, and gas injection wells and their inherent risks
Adapts easily to existing oil field water disposal schemes
Provides simple and relatively easy operation
Provides significant capital and operating cost advantages over other acid-gas recovery/disposal processes
Eliminates H₂S emissions
Reduces greenhouse gas (CO₂) emissions
Eliminates the possibility of creating pockets of gas in the aquifer, thereby reducing subsurface disposal concerns
Improves well injectivity due to ongoing acidizing effect and lowers injection pressure
Reduces hydrate formation and well bore sulfur deposition concerns associated with other schemes
Cleans injection wells with mildly acidic water, thereby reducing workover acidizing frequency
Increases possibility of oil recovery with a CO₂ effect when water is injected into the producing reservoir.

The process also has a potential for being applied outside the oil and gas industry. For example, the process might be used in mining, oil sands, and metallurgical processes, as well as with power plant waste gas streams coupled with aquifers that contain water not suitable for human consumption.

Sour gas

To meet sales specifications, many sour natural gas processing plants in Alberta remove acid gases (H₂S and CO₂) from the raw gas stream with an amine sweetening process.

In this process, the raw inlet gas is contacted with an amine solution (MEA, DEA, or MDEA) that absorbs H₂S and CO₂ from the gas stream. The amine solution is subsequently regenerated by steam stripping that releases H₂S and CO₂.

In large plants, the recovered H₂S can be economically recovered and converted to sulfur. However, in smaller plants, the conventional practice has been to flare the recovered acid gases to the atmosphere.

Flared acid gas emits SO₂ and CO₂. Both gases are considered detrimental to the atmosphere.

In early 1993, Alberta's former Energy Resources Conservation Board (ERCB) began requesting operators in the Provost area to investigate alternative acid-gas disposal methods. This request was consistent with the desire to curtail acid-gas flaring in Alberta to eliminate atmospheric emissions.

The ERCB's approval of another party's application for acid-gas disposal by subsurface injection in the gaseous phase prompted PanCanadian, as a mineral interest owner offsetting the project area, to conduct a detailed technical study of this scheme.

While recognizing the proposal's objectives, PanCanadian's study raised the concern that acid gas is not readily dissolved in formation water and possibly formed free acid-gas pockets in the aquifer. These acid-gas pockets created the following two concerns:

The acid gas could migrate easily to indeterminant locations in the reservoir because the gas has high mobility.
For new wells intersecting these gas pockets, an unexpected or abnormal situation would be created with the potential for a critical sour well drilling environment.

Because of these concerns, PanCanadian developed an alternative scheme that mixes and dissolves the acid gas in formation water at the surface. The new scheme would eliminate gas pocket concerns around the well bore, and inherently reduce acid-gas disposal costs.

Process description

PanCanadian's process is employed in conjunction with oil production and treating facilities.

Sour oil/water emulsion is produced to an oil battery at which oil and water are separated and sent to tanks. Produced water is pumped from the storage tank to an injection or disposal well and then down to a subsurface formation.

A gas conservation plant treats the sour gas to produce sweet gas and natural gas liquids. In the gas-treating process, acid gases are removed with a conventional amine system. Acid-gas disposal is achieved by dissolving it in the produced water at elevated pressure and injecting the combined acid gas and produced water stream into a subsurface formation.

Published data indicate that H₂S and CO₂ solubility in water increases with increasing pressure, decreases with increasing temperature, and decreases with increasing salinity.

The simplest of these parameters to control in the field is pressure. Therefore, the minimum mixing pressure required for complete acid-gas dissolution in the water depends on the mixing temperature, quantity of acid gas disposed of, and amount of water available.

Fig. 1 [49,103 bytes] is a typical solubility curve for acid gas and water. Recognizing this concept, the acid gas and water can be mixed before, after, or at any intermediate pumping stage.

With a pump downstream of mixing, it is important to completely dissolve the acid gas in the water prior to pumping the fluid, and to include sufficient head above the vapor pressure to meet the pump's net positive suction head (NPSH) requirements. This will avoid vapor locking and subsequent pump damage.

After evaluating several mixing methods, PanCanadian selected fixed vane or static mixing elements. The mixing elements shear the gas stream into micron-sized bubbles to allow quick dissolution into the water phase. Once the fluid is sufficiently mixed, it can be pumped and disposed of through a conventional disposal system for sour produced water.

The process simplicity makes this disposal scheme attractive.

Two process variations have been installed. The first, near Provost, Alta., is at the PanCanadian David Battery No. 3 (Fig. 2) and the second at the PanCanadian Thompson Lake facility, near Hardisty, Alta. (Fig. 3).

Fig. 4 [84,076 bytes]shows a schematic flow diagram of both acid-gas-disposal facilities.

At the David Battery, a single-stage centrifugal booster pumps water from the stock tank to 200 psig. The acid gas is drawn off the amine plant at about 5 psig and compressed through a screw compressor. The two streams are blended and mixed through 16 static mixing elements.

The aqueous mixture continues to the main multistage centrifugal charge pump with a discharge pressure of 1,000 psig. This high-pressure fluid is delivered to two disposal wells, about 1,600 ft away, through an existing pipeline. The fluid is injected down a standard internally coated tubing string.

The David facility handles about 126,000 b/d of produced water. However, only 44,000 b/d are used to mix with the 500 Mscfd of acid gas at a 2:1 acid gas to water volume ratio.

A new booster pump was installed to provide higher discharge head to facilitate a higher intermediate mixing pressure.

The process at Thompson Lake involves compressing the acid gas to 1,000 psig prior to mixing with water. The produced water pump capacity is about 44,000 b/d and the design acid gas rate is 2,663 Mscfd. A higher mixing pressure is required because of the 10.7:1 ratio of gas to be absorbed to the volume of water available.

To achieve the higher mixing pressure, a reciprocating compressor designed for a wet and severely sour gas application was selected. The acid-gas compression was designed without dehydration, as has been the practice for some acid-gas injection schemes.

Fig. 5 shows the fluid phase envelope and the compression cycle encountered with this compressor. Note that the final compression stage approaches the dense phase region at design temperatures. In addition, mixing in the aqueous phase at surface gives the advantage of reduced injection pressures because of the hydrostatic head available in the well tubing.

The resultant pH of the acid-gas-laden produced waters was calculated to be 5.3 for the David scheme and 4.7 for the Thompson Lake scheme using the Solmineq geochemical modeling program. Solmineq is a trademark of Alberta Research Council, Edmonton.

Fig. 5 [51,732 bytes] These conditions are within the capability of conventional materials typically employed in the industry for handling saline produced waters.

Process testing

CO₂ solubility in water is well known and is reported by Perry and Chilton.² H₂S solubility has been researched less extensively, but estimates are reported by Lee and Mather.³ However, research on the solubility of the combined CO₂/H₂S gas mixtures in water is limited.

A combination of process simulation, laboratory experimentation, and field-testing was undertaken to confirm that a complete or near-complete CO₂/H₂S gas mixture dissolution could be obtained at reasonable pressure, temperature, and volume ratio conditions.

Process simulation used the Hysim process simulator.

Fig. 6 presents the simulation results for the solubility of a 68% CO₂/32% H₂S gas mixture in water at 60° F. and 84° F. as a function of pressure.

Lab tests

Laboratory experiments were conducted at Hycal Energy Research Laboratories Ltd. to confirm the results obtained from the Hysim process simulator.

A high-pressure cylinder was half filled with water at 70 psig and 60° F. The experiments used actual produced water sampled from the David facility (Table 1 [22,448 bytes).

The remainder of the cylinder was filled with a 70% CO₂/30% H₂S gas mixture synthesized in the laboratory.

Cylinder contents were thoroughly mixed and allowed to reach equilibrium. Free vapor remaining in the cylinder was removed at a constant pressure. A liquid-phase sample was subsequently flashed to ambient conditions and the gas evolved was measured.

A second experiment at 1,050 psig and 60° F. used the same procedure. Results from these experiments compare favorably with those from the process simulation (Fig. 6 [100,728 bytes]).

Field tests

Field testing assessed the time factor in dissolving gas in produced water because simulation and laboratory testing assumed "infinite" time to obtain equilibrium conditions.

The testing determined how much gas would dissolve in about the 2 sec that it takes for the stream to flow from the injection point to the injection pump.

Testing was conducted at the David facility under full production conditions.

The test used the water charge pump, a Bingham 3 X 4 X 8 3/4B MSE-9 horizontal multistage centrifugal pump (300 hp). This pump is similar to the charge pump used in the final scheme, but smaller in capacity, rated at 110,000 b/d at 2,420 ft total developed head (TDH). This pump operates essentially in parallel to the larger 6 X 8 X 11D MSD-5 pump (1,100 hp).

The first test injected 100% CO₂ between the booster and injection pumps at varying gas flow rates. Pump parameters measured included developed head, NPSH, power consumption, casing/bearing housing vibration, and suction-nozzle fluid-borne pulsations.

The use of CO₂ was less hazardous than process acid gas, and did not require any piping modifications at that time. Because H₂S dissolves more readily in water than CO₂, testing with 100% CO₂ was expected to yield results more conservative than when an acid gas containing both H₂S and CO₂ is introduced.

CO₂ came from a portable liquid storage tank. The liquid CO₂ passed through a vaporizer that supplied CO₂ in the gaseous state at adequate pressure for injection.

An orifice meter, installed in the CO₂ supply line, measured gas flow rate. The interstage line pressure was 45 psig.

The first field test included a temporary injection header with two injection nozzles. One was upstream of the static mixer element and the other downstream to measure the effect of the mixer. Eight Koch 4-in. static mixer elements were used.

The charge pump performance was affected by introduction of small CO₂ amounts into the fluid stream. The pump's developed head began to drop immediately when the gas was introduced.

The flow rate began to surge, and vibration and pulsation measurements indicated that free gas was entering the pump, resulting in cavitation. The static mixer improved the dissolution, but the results followed the same trend with slightly more CO₂.

A second field test, after modifying the design of the system, increased the mixer elements to 16. A new booster pump increased the interstage pressure to 140 psig.

The pump was affected when gas was first introduced. However, this time the developed head of the pump only slightly decreased when gas was introduced upstream of the mixer elements.

The gas flow rate was gradually increased to a maximum of 261,500 scfd, which was 2.5 scf acid-gas/cu ft of water. Fig. 7 [46,699 bytes] shows the relationship between this ratio and the pump's developed head, expressed as a deviation from the head with no gas, for both tests.

Vibration and pulsation measurements indicated that some free gas still reached the pump. The gas at the pump inlet was less than 2% of total inlet volume (at inlet conditions) as evidenced by the relatively small deviation in the pump's developed head.

The additional mixer elements and the higher interstage pressure drastically improved results. The conclusion was that the design premise was successful.

The long-term effect of the small volume of free gas on the pump is unknown. A monitoring program on the final installed pump evaluated any long-term effects. This included a complete dismantling and inspection of the pump after a significant operating period.

Vibration and pulsation readings indicated a negligible effect, relative to normal wear and tear of a similar pump in produced water-injection service.

References

ERCB Information Letter IL 88-13, Sulfur Recovery Guidelines-Gas Processing Operations, 1988.
"CO₂ Solubility Data," Chemical Engineers' Handbook 5th Edition, Perry & Chilton.
"H₂S Solubility Data," Ber Bursenges Physical Chemistry, Lee & Mather, 1977.

Bibliography

McMullen, J., and Webster, C., "Acid Gas Disposal Schemes at PanCanadian Petroleum," Environmental Services Association of Alberta Conference on Flaring Technology, Edmonton, February 1996. Longworth, H.L., Dunn, G.C., and Semchuck, M., "Underground Disposal of Acid Gas in Alberta, Canada; Regulatory Concerns and Case Histories," Paper No. SPE 35584, Gas Technology Conference, Calgary, April 1996.
Lock, B.W., "Acid Gas Disposal, A field Perspective," Third Quarter Technical Session, Canadian Gas Processors Association, Calgary, September 1995.
Wichert, E., and Royan, T., "Sulphur Disposal By Acid Gas Injection," Canadian Gas Processors Association Meeting, Calgary, September 1995.
Ho, K.T., McMullen, J., Boyle, P., Rojek, O., Forgo, M., Beatty, T., and Longworth, H.L., "Subsurface Acid Gas Disposal Scheme in Wayne - Rosedale, Alberta," SPE, Alberta Energy and Utilities Booard, Spring 1966.
Beard, T.L., "Acid Gas Absorption Process," GPA meeting, Midland, Tex., May 2, 1996.

The Authors

Dave Kopperson is a specialist in corrosion and chemicals in the operations engineering department of PanCanadian Petroleum Ltd., Calgary. His responsibilities include integrity and reliability of pipelines and pressure vessels as well as design, implementation, and monitoring of chemical programs.
Kopperson has a BS in chemistry from the University of Waterloo. He is a member of NACE.

Steve Horne is a project manager for business development at Novagas Canada Ltd., Calgary. He previously was with PanCanadian. He is involved in the design, construction, and operation of natural gas processing facilities, liquids extraction, gathering infrastructure, and acid gas injection facilities. Horne has a BS in chemical engineering from the University of Alberta. He is a member of the Canadian Gas Processing Association (CGPA) and is a registered engineer in Alberta.

George Kohn is a senior engineer, maintenance, at PanCanadian in Calgary. His work involves rotating equipment engineering. Kohn has a BS in mechanical engineering from the University of Alberta. He is a member of ASME, ASM, Vibration Institute, and CGPA.

Dwayne Romansky is coordinator-production engineering in PanCanadian's south central Alberta business unit. He is responsible for all aspects of production engineering and provides technical guidance to staff.
Romansky has a BS in petroleum engineering from the University of Alberta.

Chong Chan is a development engineering coordinator in the Palliser business unit of PanCanadian, Calgary. His experience is mainly in the areas of reservoir management including exploration, reservoir development, and exploitation.

Gordon L. Duckworth is co-owner of DPH Engineering Inc. in Calgary. He is involved in the design and project engineering of oil and gas processing facilities. Duckworth has a BS in chemical engineering from the University of Alberta. He is a member of the Association of Professional Engineers (Alberta and Saskatchewan), and the CGPA.