You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.
To ensure continued access to JPT's content, please Sign In, JOIN SPE, or Subscribe to JPT
Eni began producing oil reserves from the Aquila reservoir in the Adriatic Sea soon after its discovery in the early 1980s. As primary production decreased, a decision was made to begin enhanced recovery with artificial gas lift. With the play in deep water (815 m) and 46 km off the southern coast of Italy, a floating production, storage, and offloading (FPSO) vessel was needed. After a 5-year run, the Firenze halted operations in 2018 because of low oil production. The complete paper examines the decision to use hydrogen-sulfide- (H2S) removal technology, the cost of operation, and the unit’s availability over its lifetime.
Introduction
As the industry searches for reserves in ever-deeper formations, the requirement of contending with sulfur increases. Several H2S-removal technologies are available, including nonregenerative liquid scavengers (triazine-based), nonregenerative solid-bed absorbents, and the regenerative liquid-reduction/oxidation (redox) process. These technologies remove sulfur from associated gas streams and do not release them to the environment. The nonregenerative technologies are often referred to as scavengers.
Process Evaluation
During the initial design phase, several H2S-removal technologies were evaluated per the following criteria:
- Turndown capability
- H2S-removal efficiency
- Degree of operator involvement required
- Maintenance requirement and waste material produced
- Proven reliability in marine conditions
The evaluation led to the selection of the liquid-redox process after it received the highest marks in four of the five criteria. Both liquid and solid H2S scavengers were considered as alternatives to the liquid-redox process for this installation. The considerations for each criterion are detailed in the complete paper.
FPSO Firenze
Comart worked with Merichem to design the H2S-removal single-lift module for the Aquila Phase II project. The Firenze (Fig. 1) is 810 ft long by 140 ft wide. The topsides can process 10,500 B/D of crude oil and 7.5 MMScf/D of associated gas. The hull has a crude oil storage capacity of 750,000 bbl. In Fig. 1, the arrow indicates the liquid-redox unit.
In addition to the six main modules, the topsides included modules for black-start compression, stripping gas compression, fuel-gas metering, fuel-gas compression (to the power turbine), steam boilers, and main power generation. Each module varied in weight from 10 to 530 tons and was designed as a single lift for topsides installation.
The Liquid-Redox Process
In this FPSO application, the liquid-redox unit processes sour-gas streams from the oil stabilizer, the sour water stripper, and, if required, the gas-lift compressor. The original design conditions included the following:
- 5.1 MMScf/D
- H2S in=11,000 ppm
- H2S out=less than 100 ppm
- 2.3 metric tons per day (MTPD)
- Sulfur pressure=73 psia
The combined sour-gas streams pass through a coalescing filter to remove any entrained liquids before entering the unit. The feed gas is then routed to a liquid full absorber (LFA), where it contacts a proprietary aqueous solution of chelated iron. H2S is absorbed into the water and then reacts with the iron to form solid elemental sulfur. The sweet gas exits the LFA and passes through a knockout pot with water wash to recover any entrained solution before entering the fuel-gas system. The solution leaving the bottom of the LFA is pumped to the oxidizer for regeneration. In the oxidizer, air is sparged through the solution and the ferrous iron is oxidized to the active ferric state.
The air exiting the oxidizer is routed to a knockout pot to recover any entrained solution. This stream contains no H2S and is routed directly to the atmosphere. Most of the regenerated solution is circulated back to the LFA. By adding reactions detailed in the complete paper for the LFA and oxidizer, the overall reaction becomes the direct oxidation of H2S to elemental sulfur. This reaction takes place at near-ambient conditions in the aqueous phase, making the chemistry inherently safer than that seen in a fired process such as a modified Claus unit.
Removal of sulfur from the unit is accomplished by directing a slipstream of circulating solution to a pressure belt filter. The filter uses high-pressure air and wash water to recover solution and produces a filter cake consisting of 65 wt% sulfur and 35 wt% dilute solution (mostly water). Consequently, for every pound of sulfur removed from the gas stream, approximately 1.5 lbm of sulfur cake are produced. This compares favorably with the approximately 10 lbm of spent solid scavenger produced for every lbm of sulfur removed. Although the sulfur cake contains 35 wt% water, the cake is dry to the touch and there is no liquid leakage from the cake. The sulfur cake is collected in 1-ton sacks and stored on deck until the next supply boat arrives.
Special Design Considerations
In addition to standard marine specifications, an FPSO requires special design considerations because of the constant movement of the vessel caused by wave motion. This movement is extremely important to consider when controlling liquid levels in vessels. Computational fluid dynamic (CFD) modeling was used to design baffling systems within the LFA and the oxidizer vessel that minimize liquid sloshing. These CFD modeling protocols enable calculations for any set of motion parameters when designing new units.
The movement of the vessel also affected standard analytical procedures used for onshore operations. Weigh scales could not be used under these conditions, posing a significant analytical problem. Volume-based techniques were developed to determine the amount of sulfur contained in the circulating solution.
Space and weight limitations on an FPSO affect equipment sizing and module layout. The H2S-removal module was constructed for a single lift onto the FPSO topside. To conserve space, the module was designed in a stacked fashion, building up rather than out. Because the system contained 141 tonnes of solution (mostly water), a considerable amount of structural steel was required.
Operating Results
After experiencing typical startup difficulties such as power outages, the liquid-redox unit has performed well. The unit exhibited its turndown capabilities and process flexibility quickly. The sour-gas flow rate varied from 1.5 to 5.1 MMScf/D, inlet H2S concentrations varied from 8,000 to 15,000 ppm, and sulfur production rates have exceeded the design capacity (2.3 MTPD) of the unit. Despite these varying gas flows and sulfur-concentration regimes, the unit consistently met the outlet H2S specification.
Modifying liquid-redox technology for functionality in an FPSO environment has not affected unit performance negatively. Since going online in January 2013, the unit has required approximately 25 hours per week of operator attention, or approximately 3 hours per day, on the continuously manned facility. The unit consistently exceeds the H2S-removal requirement (typically less than 20 ppm in the treated gas) and has achieved all performance guarantees and operational requirements. Because of the value of the sulfur cake as a fertilizer and fungicide in Italian vineyards, the operator disposed of its sulfur byproduct onshore at no cost.
Availability
In the past, reliability of liquid-redox processes has suffered because of solid sulfur falling out of solution and plugging piping and equipment. The author’s company has adopted design and operating practices that prevent sulfur from settling in the wrong places within the unit. In addition to strict adherence to velocities in design of the unit, the main operational method is to use air blasts or water injection placed strategically throughout the unit in regions of low flow. Nozzles send bursts of air or water into stagnant areas, preventing sulfur buildup.
In 2015, the amount of sulfur to be removed began to exceed the design amount of 2.3 MTPD. This occurred over production periods from days to weeks. The sulfur belt filter was unable to keep up with the added load, causing the solution to carry more solids than the design stipulated. The result was more-frequent outages to clear plugging in some piping. Operators were able to clear these sections of pipe in just a few hours.
Availability of the unit on the Firenze has ranged from 97 to 98% onstream. No shutdown for maintenance occurred in the first 18 months of operation. The operator then moved to a yearly turnaround schedule. The annual turnaround takes just 4–5 days from gas-off to gas-in. The remaining time offline consisted of infrequent, short outages when sulfur removal was high.
Cost of Operation
The two largest operating-cost components for a liquid-redox system are chemical consumption and electrical usage. The electrical demand is constant, even with changing sulfur load. Electricity is generated onsite. Because the gas to fire the turbine would otherwise be injected as artificial gas lift for the reservoir, fuel cost is near zero. The major operating cost per pound of sulfur removed was $0.29. Minor costs included 3 hours per day of operator time to conduct solution testing and other tasks.