Last week, we presented a problem with regard to “peek-a-boo” SCVs, where data results are inconsistent at best. Because of this variability in the 'show'ers, the standard test may not capture the event appropriately. As prefaced in last week’s post, your data shows periodic flow and no buildup pressure. The well will fail the bubble test, depending on when it was tested. The $64,000 question is, is the well good or not?
Once you have your measurements, what’s next?
The next step is to determine where the flow is coming from. This is where things get tricky. It’s not as if you can sit with a bottle or bag to try and collect a sample. For some wells, it could take days to get enough sample. Fortunately, we have developed a couple different passive sampling systems to do just that. Our sampling system 'sits' on the well and collects a representative gas and/or liquid sample so the matrix is protected from environmental degradation. This allows the highest-quality sample collection, providing the best opportunity to identify the source of the flows. We have developed a couple types of samples to deal with different scenarios, the two main scenarios being passive samplers on regular wells, and break-out samplers for steam injection wells.
Without getting into the technical parts of the samplers, the main idea here is to collect a representative sample from the SCV whereby the sample is protected from environmental degradation. What good is a sample if it doesn't represent the fluids coming from the SCV, because they have degraded during collection prior to analysis? With a well-collected sample, the fluids can be sent for chemistry and isotopic analysis, on which we can then use the results to identify from where the fluids are coming. Some engineers just like to 'grab' a sample, taking sampling for granted, incorrectly believing that any sample will do. We have extensive data that illustrates what happens to samples when they’re not properly collected, and the results can lead to the wrong conclusions.
Making sampling a priority, and doing it correctly, will make the most of your data. It will pay off in the long run.
