TECHNOLOGY UPDATE

Modeling H2S-Scavenger Efficiency Cuts Costs

The problems associated with produc- JIP work complements some extensive mately 15% are feasible with a high
tion of H2S are well defined within optimization work on a Shell field, probability of success through imple-
the oil industry, with the major issues where practical considerations to the mentation of the following:
related to safety, corrosion integrity, chemical application had a proven sig- • Improved monitoring and mea-
and production deferment. Shell now nificant effect on chemical use and sub- surement of H2S in the export gas,
has at its disposal a software model sequent reduction. This practical test which would allow optimization and
to predict H2S-scavenger efficiency in work offshore clearly defined the many control of the scavenger-dosage regime
an effort to reduce costs and optimize and varied considerations to the appli- by up to 15%
chemical treatment regimes. The com- cation technique required to achieve • Improved delivery of chemical by
pany also has available a test flow loop success. The results from one example priming the pulsation dampener and
for evaluating scavenger performance are shown in Table 1. using an appropriately sized atomiza-
under a wide range of simulated Similar reductions in chemical use tion nozzle
field conditions. were noted for two other platforms • Maximized contact between the
Phase 2 of a joint-industry proj- in the field during this extensive off- injected chemical scavenger and the
ect (JIP) with the Chicago-based Gas shore optimization study. The main fac- sour gas from the low-pressure com-
Technology Institute (GTI), which tor contributing to this reduction was pressor, through use of an additional
Shell recently completed, provided the the relocation of the injection point. scavenger-injection point at the high-
basis for the company’s advances. The Further cost reductions of approxi- pressure separator discharge

TABLE 1—H2S SCAVENGER USE OPTIMIZATION RESULTS

Pre Post
LP discharge LP discharge HP HP
Operating parameter Units discharge discharge
Case 1 - Min Case 2 - Max Case 1 - Min Case 2 - Max
consumption consumption consumption consumption

Gas-flow rate million m3/d 6.01 6.01 6.01 6.01

Pressure bar 60 60 22.4 22.4
Temperature °C 71 71 29 29
H2S initial in gas phase ppm 4.3 4.3 4.3 4.3
H2S final ppm 1.5 1.5 1.5 1.5
Pipe diameter - if Direct injection in. 24 24 28 28
Pipe length m 20 20 40 40
H2S partial pressure psi 0.0046 0.0046 0.0023 0.0023
Scavenger used water-soluble HSW82165 HSW82165 HSW82165 HSW82165
Actual gas rate m3/d 9.44E+0.4 9.44E+0.4 1.67E+0.5 1.67E+0.5
H2S to be removed kg/d 24.50 24.50 24.50 24.50
Actual scavenger consumption L/d 975 1351 605 740
Scavenging efficiency L/ppm H2S rem/million Nm3 57.94 80.28 35.95 43.97
Volume-to-surface ratio ft3/sec/ft2/ft 6.14 6.14 9.33 9.33
Velocity m/s 3.75 3.75 4.88 4.88
Ratio of V/Vc 1.53 1.53 1.41 1.41
Transit time seconds 5.34 5.34 8.20 8.20
Average scavenger consumption L/d 1163 673
Average treatment rate HSW82165 L/ppm H2S rem/million Nm3 69.11 39.96
N=Normal LP=Low-Pressure HP=High-Pressure

26 JPT • DECEMBER 2007

Technology Update

Fig. 1—A schematic and photograph of the scavenger test loop at the
Gas Technology Institute are shown.

• Use of a pump suitable for the volu- Adding Value to the

metric injection rates Chemical Contract
• Fitting the flowing system with a On the basis of the optimization stud-
retrievable insertion probe for ease of ies, the improved prediction of scav-
maintenance and reduced downtime enger efficiency can add value to an
Additional savings of more than operator’s chemical contract from sev-
10% in chemical costs are achievable eral standpoints:
through automating the injection, with • Environmental
a payout period of less than 1 year pro- ° Better compliance with the United
jected for this. Kingdom’s Petroleum Operations Notice
Implementing the above recommen- 15D discharge limit through reduced
dations will yield a positive environ- chemical use
mental impact by achieving up to a ° Reduced tank movements, cor-
30% reduction in the concentration of responding to reductions of approxi-
unspent scavenger in the water that is mately 50 per annum at each of the
discharged overboard. platforms in the studies

JPT • DECEMBER 2007

Technology Update

• Safety communication on these issues given a against hundreds of data sets from the
° Reduced road transport higher profile. GTI test loop, as well as from several
° Reduced tank filling and decanting commercial H2S-scavenging facilities.
° Reduced crane and forklift lifts Computer-Model Development A schematic and photograph of the
° Reduced shipping During Phase 2, a new steady-state GTI test loop are shown in Fig. 1. The
• Cost mechanistic model was implemented, pipe is made of ¾-in. 10 CS schedule-80
° Transit tank hire based on more-rigorous modeling of pipe with total contact length of 240 ft.
° Shipping tanks the two-phase flow and mass transfer. The entire structure is housed inside a
° Chemical costs The underlying equations developed 20×12-ft fume hood equipped with a
were based on both theoretical and carbon adsorption bed. This unit has
Additional Benefits experimental work sponsored by GTI a maximum flow of 400 Mscf/D of gas
to Platform Operations and participant members over the past under both once-through and recycled
In addition to the above benefits, one of 10 years. The model is programmed in flow conditions at higher temperatures
the platforms studied now has an opti- Microsoft Visual Basic Software and has (function of gas-flow rate) and at pres-
mized maximum dosage rate for total a graphical user interface. The calcula- sures up to 1,200 psig. Testing in the
production vs. H2S volume produced. tion method is based on determining scavenging loop can also be used to
Information on each producing well’s the mass-transfer rate of H2S from the determine if proposed scavengers are
H2S loading is constantly updated. This gas to the liquid at each point along the sufficiently reactive to be capable of
has enabled the platform to set dosage flow path inside the length of the pipe removing the H2S to the desired levels in
rate on the basis of anticipated H2S, or heat exchanger being modeled. the required pipe lengths. It can be used
rather than relying on gas-export read- The rate of H2S absorption per unit to compare the performance of scav-
ings and then making manual pump length is expressed as a function of engers against each other and against
adjustments to set dosage. the local conditions at that point. The known scavengers, to determine the
Cost and use levels of H2S scavenger equation is then solved by use of a likelihood of achieving the performance
now receive a much higher priority on standard fourth-order Runge-Kutta specifications with each scavenger.
the platform, with platform-to-shore method. The model has been validated The activation energy, diffusion,
viscosity, surface tension, and other
mass-transfer phenomena, however,
are dependent on the solvent (i.e.,
the scavenger being used). The model
will be expanded and improved signifi-
cantly during the Phase-3 effort. Several
new features will be incorporated:
• Temperature-dependent chemical
kinetics for different scavengers
• Vertical-upflow section calculations
• Diffusion effects
• A different set of flow-regime char-
acteristics
These program features will increase
the accuracy and robustness of the
model to handle a wider range of oper-
ating conditions.
During Phase 2 of the JIP, experimen-
tal techniques were developed in the
GTI vapor/liquid-equilibrium unit to
measure kinetics along with the effects
of mass transfer. A schematic and pho-
tograph of this unit are shown in Fig. 2.
The kinetic data will be used to develop
a more temperature-dependent kinetic
model for each scavenger. In Phase 2,
it was limited to a single scavenger and
under a limited range of conditions.
In Phase 3, this will be broadened to
include scavengers of interest over a
wide range of conditions.
Pseudofirst-order kinetics that was
observed in Phase 2 for ethanol-based

28 JPT • DECEMBER 2007

Fig. 2—A schematic and photograph of the vapor/liquid-equilibrium unit
at the Gas Technology Institute are shown.

triazine should apply across the entire H2S in each scavenger. These data will
range of ethanol-based H2S-scavenger be used to develop a liquid-mass-trans-
products. If deviations are encountered, fer coefficient based on the scavenger
the product is considered suspect and of interest.
could be contaminated. The level of
residual intermediates and active etha- Looking Forward
nol/triazine purity will be considered in The project will continue to evalu-
the work. ate the model performance as changes
The temperature dependence will be are made. The model predictions will
sought for each scavenger, with the be compared with experimental data
understanding that diffusion interfer- from the laboratory and field-test data
ence by the surrounding package (i.e., to quantify the prediction errors and
the type of surfactant added and/or the identify operating regimes or physical
winterization additives that may be conditions where modeling difficulties
required) might cause a deviation of arise. The continual evaluation and val-
the observed activation energy from the idation of the model provide a measure
predicted value with such additives. of quality control and confidence, and
The current model uses empirical a degree of safety factor, that is required
relations for the diffusivity of H2S in for effective application. JPT
water. Experiments will be set up to Information provided
determine the gas/liquid diffusivity of by Calum Buchan, Shell

JPT • DECEMBER 2007