Validation of Salt Water Disposal Well Treatment Plants with Handheld Oil in Water Sample Analyzer
Salt Water Disposal (SWD) well operators often dispose a substantial volume of oil inadvertently when injecting salt water, thereby losing the opportunity to sell the oil on the market and potentially reducing the ability to inject water in the well. Injection facilities are typically operated with minimal instrumentation or system optimization, which means the operators don't know how much, or when, oil is being disposed.
A typical oil field tank battery for oil/water treatment consists of a collection tank, a gun barrel tank, and storage tanks for water and oil. The purpose of these tanks is to produce clean oil and water and remove solids in the process. The reality, however, is that often the water has a significant amount of oil and solids that do not separate. Some of the oil is emulsified, and some is coating the solids, sinking to the bottom of the tanks.
The capacity of a typical gun barrel tank is between 750 and 1000 barrels and can treat up to 10,000 barrels per day. At that rate water is retained in the tank for about 2.4 hours, allowing the oil to separate and be skimmed.
Typical gun barrel tank
At a truck unloading facility, the water flow rate can vary wildly as trucks dump water at irregular intervals. In pipeline fed facilities, the flow rate can be more stable, but can still have quick changes.
A typical oil/water separation system is a relatively simple system with no, or minimal, moving parts and is designed for a set of conditions (flow rate, oil concentration, oil gravity, pH, solids content, etc.). When the process is operating within the design parameters, the water quality will be good, the injection well will operate with low maintenance costs, and most of the available oil will be collected. However, in processes that fluctuate greatly, it is common for the conditions to be outside the design parameters. Also, as these SWD facilities grow with increased water volumes, the treatment systems will be even more challenged by upset conditions.
Surveys and Pilot Demonstrations
Before approving a facility improvement project, the owners of a SWD well facility must have some data on current process conditions and performance. Usually a pilot, or demonstration, plant is tested before a full-scale facility is built. Before and after flow rate and oil concentration data should be collected from these systems to verify the design is correct and the capital expenditure can be approved.
Ideally, a continuous online monitor would be used for collecting data to determine upset frequencies, duration and normal conditions. Many of the SWD facilities and demonstration plants do not have a budget that allows for continuous monitoring. However, high frequency sample analysis can be performed on site, using a portable analyzer. Easy to use, low cost field analyzers, such as the TD-500D, are essential for optimizing the effectiveness of treatment systems during pilot studies and meeting a budget that includes frequent measurements.
Two Black wallProcess Studies
In a recent Blackwall Process demonstration, preliminary data reported that the SWD facility treated 4,600 barrels of water per, and that the gun barrel tanks skimmed an average of 380 barrels of oil per month. Oil in water measurements averaged 2,753 parts per million (ppm).
Backwall Process installed its water treatment system for a 2-week trial. During this trial the system treated 2,000 barrels of water per day and 570 barrels of oil were skimmed. To compare this to the original flow rate, approximately 2,622 barrels of oil would have been skimmed per month. This skim volume shows that the water averaged 20,357 ppm oil content. It also proves that the original treatment system was only removing about 10 percent of the oil, with the remainder being lost in the disposal well. This data also illustrates that the disposal well probably had significant maintenance costs due to the high oil concentration.
This demonstration facility showed significant oil removal, but was not able to collect frequent enough data during the trial to analyze upset frequencies or upset magnitude.