Oct-2016

Are produced water emission factors accurate?

Rules of thumb are often used for estimating emissions from Produced Water storage tanks, especially when entrained hydrocarbons are present.

Kyle Ross
Bryan Research & Engineering

Article Summary

The rules of thumb are used because of difficulties in obtaining accurate compositional information as well as deficiencies in the estimation calculation methods. The different rules of thumb yield substantially different emissions estimates, and it is questionable whether some provide accurate estimates. A new process simulator-based method is presented that overcomes some of the sample analysis and calculation shortcomings. Predictions of the new method are presented that show the effect of hydrocarbon entrainment on produced water emissions. In addition, definitions and emissions estimates of some of the currently used rules of thumb are presented and contrasted with the new method.

Introduction
Rules of thumb are often required to estimate hydrocarbon emissions from produced water storage tanks due to lack of sampling or inadequacies of sample analysis, the “1% rule” being the most common. The premise behind the 1% rule is that entrainment from upstream separation introduces hydrocarbon liquids into the produced water tank. This entrained material forms a layer of hydrocarbons that float on top of the water in the tank and should be expected to increase total emissions. As it is difficult to measure the entrained oil content in the water fed to the tank, there is uncertainty in how much of this entrained oil is lost to emissions. The 1% rule is therefore applied as an estimate.

If 1% entrainment is assumed, a problem arises as to how this should be incorporated into typical emissions calculations. In discussions with industry colleagues, this author has found at least four different definitions for the 1% rule and how it should be applied. As each method produces significantly different estimates, the question then arises, do any of these methods provide accurate estimates of produced water tank emissions? Which of these methods makes the most sense?

Storage tank emissions are commonly divided into four categories; flashing, working, breathing, and loading. Flashing losses occur from vaporisation of components in the tank inlet due to a pressure decrease and/or temperature increase of the material. These losses occur when the material is introduced to the tank. Working, Breathing, and Loading Losses (WB&L) occur as the material in the tank weathers or as it is removed. They are all caused by a changing vapour space in the tank. For working losses, this is by liquid level changes in the tank. Working losses increase as tank throughput and hydrocarbon vapour pressure increase. Breathing losses are the result of daily ambient temperature changes as the changing temperature causes the vapour space in the tank to expand and contract. Breathing losses increase with vapour pressure and are not directly influenced by tank throughput. Loading losses occur as vapors in a cargo transport vessel are displaced by liquid being loaded into the vessel. These losses increase as tank throughput and vapour pressure increase.

Before evaluating the various methods of applying the 1% rule, it makes sense to discuss a few items. How do hydrocarbons make their way to the produced water storage tank? Is there a logical method that would provide a more rigorous estimate of emissions than methods currently employed?

A diagram of a typical well site configuration can be seen in Figure 1. After exiting the well head choke, the well stream is commonly a three-phase mixture of gas, hydrocarbon liquid, and aqueous liquid. The gas is first separated from the liquid mixture in a high-pressure separator (HPS). After leaving the HPS, liquids will then flow to a heater treater (HT), where more gas is removed, and the two liquid phases are separated at an elevated temperature and reduced pressure (typically 20 to 50 psig). The produced water leaving the HT then flows to the Produced Water Storage Tank, and the hydrocarbon stream to the Hydrocarbon Storage Tank where they are stored until loaded and transported away from the wellsite.

Hydrocarbons can make their way to the produced water storage tank in two ways, either by dissolving in the water or by mechanical carry under known as entrainment. While it is often said “oil and water don’t mix”, hydrocarbons are in fact slightly soluble in water, with lighter hydrocarbons and aromatic components being the most soluble. This solubility increases somewhat proportionally with pressure, and it is dependent on the concentration of salts in the produced water. The solubility is highest in pure water and declines with increasing salt concentration. An estimation of dissolved hydrocarbons on a “salt-free” basis would produce a conservative estimate if using a process simulator. There are also analytical techniques that mimic the pressure reduction from upstream separator conditions to storage conditions, providing an additional estimate of the flashing emissions from dissolved hydrocarbons in the produced water.

The results of these estimation techniques will confirm the expected low solubility of hydrocarbons in the water, and therefore a low contribution of overall emissions by the dissolved hydrocarbons. Table 1 gives the separator conditions and compositions for four samples of pressurised liquids leaving HTs. For Sample 1, the equilibrium dissolved hydrocarbon solubility in pure water predicted by ProMax® [1] is below 100 ppm, indicating a low contribution to overall emissions from dissolved hydrocarbons.

The second way that hydrocarbons enter the produced water storage tank is as an entrained second phase, and this second phase is a significantly greater contributor to emissions than the dissolved hydrocarbons. To estimate emissions from the entrained liquid, the composition and volume of the material must be known. The entrained hydrocarbon droplets are the same material as the hydrocarbon liquid stream leaving the HT, a material of which the composition is typically known. However, what is usually unknown is the size or total volume of the droplets that are entrained.

Separators are designed to remove droplets of a certain diameter and larger. Typical minimum droplet diameters are in the 100- to 150-micron range. A fraction of the droplets smaller than this threshold will be entrained due to inadequate residence time in the separator. Directly measuring these droplets, either in terms of size or total volume, is difficult. The amount of entrainment is therefore typically assumed or estimated. Some individuals assume 200 ppm hydrocarbon leaving the HT with the produced water. A separator manufacturer questioned for this paper stated their typical design point is 0.5% (5000 ppm). Meanwhile, companies reclaiming hydrocarbons from produced water estimate 1 to 2% of the material entering the tank is entrained hydrocarbons.

With such widely varying estimates of entrainment and no reliable or cost effective way to measure the hydrocarbon content, it is helpful to look for alternative ways to estimate entrainment. Fortunately, one can look at what is leaving the tank to estimate what is entering. As the produced water with the entrained hydrocarbon droplets enters the storage tank, a portion of the hydrocarbons will flash due to the lower pressure and enter the vapour space. The remaining liquid hydrocarbons will eventually coalesce as a separate layer on top of the water. Care is taken to avoid removing material from the hydrocarbon layer when unloading the water. The hydrocarbon layer itself is removed periodically by the operator in known quantities. As the hydrocarbon layer is only removed intentionally, the rate at which entrained hydrocarbons enter the tank can be estimated by monitoring the rate at which they are removed.

The Environmental Protection Agency (EPA) has recommended that Chapter 7 of AP 42 [2] be followed for estimating tank WB&L emissions for organic liquid storage vessels. The rules of thumb discussed in this paper all rely on AP 42 in some manner.