Estimating ethylene glycol injection rate for hydrate inhibition

In refrigeration and choke plants, gas is cooled to low temperatures (usually between -10 C and -30 C), to meet hydrocarbon dewpoint specifications and/or recover valuable liquids. Ethylene glycol is typically injected in the facilities to ensure that hydrates do not form at these low temperatures.

Process engineers are commonly required to estimate the necessary ethylene glycol (EG) injection rate to inhibit hydrate formation. Process simulators such as Aspen HYSYSTM can be used to estimate the EG requirement.

Before discussing the simulator methods, it is useful to understand the alternative chart/empirical methods which can be used. The Gas Processors Suppliers Association (GPSA) handbook is a good resource for these methods, which require the engineer to:

1. Estimate the hydrate formation temperature using an appropriate method. For sweet gas, the Katz method1 can be used to predict hydrate formation temperature. The Baillie & Wichert method2 can be used to predict hydrate formation temperature in high H2S-content gases.
2. Calculate the inihibitor requirement. The Hammerschmidt equation3 can be used to calculate the required concentration (and therefore flowrate) of ethylene glycol or other inhibitors in order to suppress hydrate formation at the temperature of interest.

Process simulators make this job easier, but it is important to know that very different results can be obtained vs. the chart/empirical methods which have been used over the years. Oil and gas companies often have their own internal expertise and guidelines for hydrate inhibition calculations and these should be reviewed, if available.

Also, for those who are not as experienced in simulation, there are many ways in which this seemingly simple calculation can be mismanaged. Here are a few tips:

• The degree of water saturation (or the amount of free water) will influence the hydrate inhibitor requirement. This aspect of the model should be simulated accurately. Also, if water is not specified, HYSYS assumes free water to be present.
• An Adjust operation can be used to manipulate the flowrate of inhibitor required to meet a target hydrate temperature. However, converging the Adjust can often be tricky, and the user should confirm that proper specifications are set and that it is indeed converged (seeImproving flowsheet convergence in HYSYS).
• The HYSYS hydrate utility offers several options (“Ng & Robinson” vs. “CSM”), as well as various hydrate calculation models such as “Asymmetric”, “Symmetric”, etc. Further details can be obtained from the HYSYS documentation, but in general, the defaults work best.
• The hydrate curve in the HYSYS phase envelope is plotted considering the actual amount of water in the stream. If there is no water specified, then free water is assumed to be present.
• The effect of salt (brine) is not considered in the HYSYS hydrate calculation. This is not usually a concern for refrigeration plants but it could be an issue in other scenarios (e.g., pipeline transmission).

Simulators such as HYSYS can be used to estimate inhibitor requirements, making it significantly easier than classic chart/empirical methods. However, care and proper engineering diligence should be used when calculating these requirements.

References

1. Katz, D. L., “Prediction of Conditions for Hydrate Formation in Natural Gases,” Trans. AIME Vol. 160, 1945, p. 140.
2. Baillie, C., and Wichert, E., “Chart Gives Hydrate Formation Temperature for Natural Gas,” O&GJ, Vol. 85, No. 14, April 6, 1987, p. 37.
3. Hammerschmidt, E. G., “Formation of Gas Hydrates in Natural Gas Transmission Lines,” Ind. Eng. Chem., Vol. 26, 1934, p. 851.

Latest articles

• Compressor blowdown calculations: Comparison of HYSYS BLOWDOWN vs the Depressuring Utility

May 13, 2013 / 22

Introduction Fluid temperatures can drop significantly during depressurization events and the h...

• An efficient approach for techno-economic evaluations of membrane technologies in SAGD Central Processing Facilities

May 13, 2013 / 1,4

Oil sands operators are continuously investigating ways to reduce operating cost, energy use, and as...

• Predictive Modeling using Machine Learning in the Upstream Oil & Gas Sector

May 13, 2013 / 20

BackgroundPredictive process models are essential engineering tools that enable better process desig...