Introduction*

Optimizing oil and gas production from wells, including lift gas injection optimization, is a common activity. Still, wells don’t operate isolated from others, and oil, gas, and water have to go somewhere… downstream into common production lines, separators, processing units, oil-from-water knockout drums, and water-from-oil knockout too, amongst many other unit operations on their way ultimately to storage or pipelines. There is limited capacity downstream, so one cannot blindly put the wells on “full” output.

As an example, wells compete within production lines. We’ve seen wells brought online, but the total production does not change downstream because the recently added well’s production pressure in the production line pushes back on all the other wells, restricting their flow.

To properly manage production, one must consider the capacities and interactions of all wells flowing into downstream operations.

Optimizing a Network of Wells

Lift gas optimization’s lift curves and production vs. FTHP (Flow Tubing Header Pressure or wellhead pressure) curves tell us we face a highly nonlinear optimization of N wells operating together as a system, some production-assisted and some free-flowing. Production-assisted wells’ production is controlled by varying the assist, such as the artificial lift gas flow rate or ESP amps or Hz. Free-flowing wells can be controlled by setting the choke on the wells. We need to determine, within constraints, what the optimal production assist and choke settings should be to deliver the total resulting liquid and gas production that the downstream operations can handle.

One optimization method that works is determining the production assist and choke setting that results in maximum production for each well independently, then “walk” the non-linear curves downward smartly to decrease theoretical production until it reaches the maximum downstream rate constraint, mindful of any total lift gas or pumping constraints.

The “walking” is done by determining which well has the least production change for the lift gas reduction, thus taking advantage of the sensitivity of production vs. lift gas, saving ample lift gas for small reductions in production. At the same time, free-flowing wells are walked up the FTHP curve, increasing the pressure as the estimated production declines. In the end, your result is a set of lift gas rates and FTHPs to operate each well to achieve a production setpoint, the downstream constraint. If free-flowing wells’ chokes are manually set, one can ask the field operations to change the choke to achieve the resulting FTHP on the gauge. If automatically set, the lift gases and FTHP (or choke positions) can be written to the DCS (Distributed Control System) as recommended setpoints. The latter is the best choice, as you can reoptimize the entire field and network in near real-time.

The curves come from multivariate nonlinear predictive models built from well test rates vs. operating conditions or multi-rate (or multi-choke) trials.

Other factors also come into play, as there may be contractual obligations by well (or even zone within a well), such as assuring at least a minimum production (a constraint on walking the well’s curves).

Results

The result of network-wide production optimization is that you can achieve maximum production of an entire asset while respecting downstream capacity constraints and production assist constraints. This maximizes total liquid and gas production and minimizes lift gas consumption simultaneously. The optimal strategies apply to any number of wells in an interconnected network, be it a handful of wells or hundreds.

The cost savings in lift gas and production assist can be substantial, and the increased production allows one to capture more of the asset’s potential revenue; on the other hand, a precise knowledge of production capacity limits is necessary so that total production remains within those limits. Production optimization for interconnected networks helps maintain a balance between the two.

A network optimization model, built with careful attention to the individual wells in the network, can enable operators to reach their production goals more easily and cost-effectively. In addition, it allows for better management of assets by providing greater insight into current performance and potential future scenarios.

Network optimization is essential for maximizing the value of an asset and optimizing the entire oil production process.

This technique has been implemented across wells in the Middle East using IntelliDynamics’ Intellect, a software suite used for modeling, predicting, and optimizing in the process industries since the year 2000.

For more information, please feel free to contact us.

* Note: The information in this article is the intellectual property of BioComp Systems, Inc.