Production Monitoring Gets Smarter With Virtual Meters
Virtual metering systems (VMS) have been around for years as a cost-effective solution for enabling reliable production metering, even in places where it was previously impossible to do any kind of well testing. It has developed into a technology that can overcome installation issues such as data unavailability, communications problems, data degradation, and system maintainability over the life of a field. By utilizing multiple methods built within the system, a VMS can predict accurate flow rates even if various sensors for measuring pressure and temperature in and around a well fail over time.
Speaking at a presentation held by the SPE Flow Assurance Technical Section, Dale Erickson outlined the role a VMS can play in production monitoring. Erickson, technical authority and technology development lead for Wood’s intelligent operations group and its automation and control business, explained the differences between VMS and physical multiphase meters, as well as the process a VMS uses to estimate flow rates.
“We started a long time ago, and we’ve basically been improving the technology as we’ve gone along. It’s been a continuous evolution where we’re making things better,” Erickson said.
Virtual metering examines the pressure drops naturally occurring in a well and, from that, infers a flow rate based on a set of models. Physical meters are either on production separators that measure conjoined flow from several wells concurrently, or on test separators that measure flow for individual wells occasionally. They can be expensive to install, and for offshore platform operators often incur additional costs for shipping a meter offshore. They are also subject to corrosion, sand, wax, and hydrate plugging, which can add on more cost for maintenance. While a VMS still requires some level of maintenance, it still puts a lesser strain on operators by being primarily software-based.
The system estimates the three-phase well flow rates in real time using the existing instrumentation within the wellbore and on the wellhead, with software based on models that extend from the reservoir to the wellhead choke. Wood’s VMS approach typical involves a reservoir depletion model, which gives a better sense of the inflow into the well. There may be an optional reservoir tank model to measure near-wellbore dynamics, along with a heat transfer model for pressures and temperatures along the tubing, choke, jumper, and manifold. Erickson singled out the choke as an important element of the modeling because it can see significant pressure drops.
“For those of you who are familiar with metering, a lot of times you have an orifice meter, and in some ways a choke is like an orifice meter. It’s the one that’s basically adjustable. This is a good means of estimating a flow rate, especially when you have a large pressure drop,” he said.
The VMS calculation procedure starts with the system collecting field data. It then filters and validates that data, first by voting in redundant measurements, then by comparing the consistency between measurements and estimating the uncertainty in the data. Treating the data often involves gross error detection and isolation using statistical tests, global tests, nodal tests, measurements tests, and unbiased estimate techniques. Reconciliation algorithms help “fix” bad data, removing outliers and helping deliver a dataset that Erickson said is closer to the true process state than raw data.
“Over the 25 years that we’ve been working on this, that’s been a big part of making virtual metering systems work,” Erickson said of the data filtering process. “These systems have to work 24/7 with no kind of human involvement in terms of fixing things. If data falls out, the system has to take care of itself.”
The system then estimates flow rates for each given change in pressure, calculates estimates and the uncertainty for each method, and then from there it can calculate the best estimated flow rate. The best estimate is a weighted average of flow rates. The overall uncertainty calculated in these estimates has two components: error propagation due to instrumentation and model uncertainty.
“If you get to the point where the choke is all the way open and there’s no pressure drop, it’s going to have a higher uncertainty and it will basically drop out of the calculation. That’s the nice thing about having these multiple methods. Depending on where you are in the flow rate range it will switch out,” Erickson said.
VMS has a number of benefits. In addition to being used in lieu of subsea multiphase meters, it has been used for surveillance and as an insurance policy for subsea metering. Erickson said the backup capabilities of the VMS have come in handy for Wood before—at one of its deepwater fields, all five of its previously installed multiphase meters had failed, but the VMS saved the company from having to invest in an intervention strategy. Because VMS facilitates improved visualization of the whole field’s production performance, users can determine if there are any biases present in an individual multiphase meter, which Erickson said can help improve the accuracy of the physical meter.
Water Survey Assesses Souring and Corrosion at Production/Injection Plant
At a water treatment plant for an onshore oil field in northern Germany, formaldehyde injection was started in 2015 as a biocide. The goal of this study was to understand the chemical parameters and microbial distribution in the water system and whether formaldehyde injection was effective.
Addressing the Gaps in Subsea Produced Water Treatment
Operators are looking for ways to better handle water coming from subsea wells, which is typically treated at topside facilities. Subsea separation systems are not equipped to discharge water back into the reservoir, so how do companies close the gaps?
Neural Networks Plus CFD Speed Up Simulation of Fluid Flow
High-fidelity 3D engineering simulations are valuable in making decisions, but they can be cost-prohibitive and require significant amounts of time to execute. The integration of deep-learning neural networks with computational fluid dynamics may help accelerate the simulation process.
Don't miss out on the latest technology delivered to your email every two weeks. Sign up for the OGF newsletter. If you are not logged in, you will receive a confirmation email that you will need to click on to confirm you want to receive the newsletter.
15 January 2020
08 January 2020
07 January 2020
09 January 2020