Mathematical Allocation verses Physics-Based Allocation? Today we are going to discuss estimating production accurately and levering your data. Estimating production rates may seem likely a trivial thing, but we have learned that many operators have put their wells on test separators only once in 15 days, or in some cases once in 2-3 months. This may be due to capacity constraints in some cases but this untimely approach can lead to delays in responses because decision making is dependent on the schedule of production tests.
Our proposed solution to this is by calculating live imputed production for improved field management with a physics-based allocation model.
There was a time when I was working in the oil field, we would feel the wellhead to understand how the well is operating. When the wellhead was warm, had a mild continuous rumbling sound, we felt comfortable that the well is producing full-bore liquid. Once the wellhead started getting warmer, we started to expect an increase in water-cut. Now that I am at OspreyData, we transitioning this understanding to an AI-based model that can be trained to infer the production of a well based on the variations in temperature and pressure with respect to their production testing conditions.
Many of us here think extensively about allocated production, which is based on a mathematical estimate. As proposed in the graphic above, we tie-up physics along with mathematics. Our model uses conditions (such as tubing, casing pressures, and temperatures) specific to the most recent and to the historical production tests to model changes in production rates for a given change in these other conditions. With this process, we are able to come up with a methodology capable of providing a live production estimate.
When considering the comparison between mathematical and physics-based allocation models, mathematical allocation is trying to come up with individual production rates for wells such that the sum of these is equal to the total production rate of a group. This may not account for a condition that could have caused a change in the production behavior of the well, hence affecting its proportionate contribution. This is different than physics-based allocation.
By bringing physics into play we are able to come up with live estimates of production potentially on an hourly basis. Such estimates can help with taking quick measures to handle frac-hits, identify gas circulation, trigger bottom-hole probes or surveys to diagnose or remediate the well.
You may ask: Does this model facilitate higher performance of event detection models? Does this make the engineer more effective at his job ?
The way I see this is that this physics-based, imputed production model is able to detect conditions such as freezing on a wellhead choke or an obstruction in tubing which causes gas circulation. There are various conditions that can cause gas circulation resulting in a drop in tubing pressure, and also a drop in tubing temperature. This should give the operator a notification that the liquid production may be down. When this happens, the operator may not know why exactly, but they are able to send out a crew to inspect the well, or be able to use our AI-based systems to get an understanding of the specific condition which caused the drop in liquid production.
Remarkably, there are still operators who do not have telemetry on their wells. If you are one of those operators, we hope this motivates you to develop your infrastructure to enable AI-based solutions for helping your wells to be operated more efficiently.
If you want to maximize your oil production and increase your ROI, contact us to set up an appointment with our sales team to discuss easy implementation on your oilfields. We say easy because Production Unified Monitoring is a cloud-based tool that can be accessed from any web browser without any hardware or software installs, meaning there are no capital expenditures. Contact Us Today!