Why are we not seeing more data-driven VFMs in operation?
In time with improved instrumentation of petroleum assets and increasing amounts of data, the buzz around AI and Machine Learning reached the petroleum industry. Is there a reason why such solutions shouldn't kick-ass in production flow modeling?
- Author
- Solution Seeker
- Publish date
- · 3 min read
Several mind-blowing achievements of data-driven solutions have been reported from other fields of research, such as AlphaGo, the computer program that beat the world champion in the game of Go. This has led to a prevalent “AI mindset” also in the oil and gas industry, with high but yet unfulfilled expectations of impact.
In recent years, there has been an increasing amount of literature on data-driven solutions to virtual flow meters (VFMs) - a soft-sensor technology able to predict the petroleum flow rates in an asset using mathematical models and already existing measurements. Some of the case-studies on real data report an error as low as 4% on predicting the oil flow rate from a petroleum well (AL-Qutami et. al. 2017). Yet, as far as we are aware, no commercial data-driven VFMs exist. Why is that?
10%
Solution Seeker has several years of experience in working with data from petroleum assets and in modeling data-driven VFMs. We believe that the traditional approaches to data-driven VFM modeling may not be robust enough for commercial use. We define a method to be sufficiently robust if it achieves less than 10% error for at least 90% of the wells modeled. To illustrate this, consider the two figures below. The figures illustrate the oil flow rate predictions from two VFMs developed for two individual petroleum production wells using an identical modeling approach. In Figure 1, the VFM achieves an error of 4%, whereas in Figure 2, the error is 45%.
![](https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=410&q=90&w=600&s=3805ead74194d366042d8b6a1575c1b8 600w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=478&q=90&w=700&s=b911325a951c3893bb40ba614471f95f 700w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=546&q=90&w=800&s=59f9b8385e11951bb8898004dd7727ea 800w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=615&q=90&w=900&s=2f8d42c0ed64a48608a4c41379d9c312 900w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=683&q=90&w=1000&s=0d879c6f4c65f598273303373d54d361 1000w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=751&q=90&w=1100&s=bca92aa574596e4aa74b17e56100120f 1100w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=820&q=90&w=1200&s=1124a7439971a9694937ee04cc367f79 1200w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=888&q=90&w=1300&s=c28b6d7fd79bddbcdc167531c7b50b44 1300w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=956&q=90&w=1400&s=e844994132c9572ca83a7bb9f670189d 1400w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1025&q=90&w=1500&s=50131d13bbe76f7b61b039e4dea3391f 1500w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1093&q=90&w=1600&s=09574c2e7968a140ffc4a5605a6163f5 1600w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1161&q=90&w=1700&s=0604dad5e588ecb61aa918c19205b86f 1700w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1230&q=90&w=1800&s=cae663521317e250591904d1b7fd37f9 1800w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1298&q=90&w=1900&s=cf088b7c28c2514b54fe79b1310511f5 1900w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1366&q=90&w=2000&s=1768d6152b581c429f3b097d047931f5 2000w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1435&q=90&w=2100&s=c34f2ff67ffddbc89b5778be5e2e2ca5 2100w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1503&q=90&w=2200&s=96ad111d211802ce9469ebaacf5c9fb7 2200w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1571&q=90&w=2300&s=851eb6d93a2cffc9c3cd237a43db5180 2300w, https://solutionseeker2021.imgix.net/images/old/figure1_2021-11-29-184959_bhki.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1639&q=90&w=2400&s=18de63a5859b79b9c02335671a4817a9 2400w)
![](https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=410&q=90&w=600&s=f85610bf7b3dffbae3cdea2417396d20 600w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=478&q=90&w=700&s=83f0b25cfa50e2b24cf18e2fc4e8ac51 700w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=546&q=90&w=800&s=1191bb2217ab73d3e9cfc6cd52559d2b 800w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=615&q=90&w=900&s=a4a4dad2c02b9bc84019435ca8077679 900w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=683&q=90&w=1000&s=6920324e44dce495607e652717db3436 1000w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=751&q=90&w=1100&s=4e4051a0169f394c24d4978fd073424f 1100w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=820&q=90&w=1200&s=125b1f80fd4b4d759e0525d928ab86dc 1200w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=888&q=90&w=1300&s=cbd65fd98a17720d8922c81b85639af4 1300w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=956&q=90&w=1400&s=782758690611516fc39f0437bb5d9738 1400w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1025&q=90&w=1500&s=7fddc1a394a6a1f55a102b6c095ea745 1500w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1093&q=90&w=1600&s=e48d5b512123d0b674619da3226a6932 1600w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1161&q=90&w=1700&s=2d556b0db71109dbb272cf9709866102 1700w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1230&q=90&w=1800&s=f3e2485342b1ca0eb2ecd070f5a2406c 1800w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1298&q=90&w=1900&s=3bbc4f5ddf8e9fe33f050c41c3b815e1 1900w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1366&q=90&w=2000&s=6505d571f18bf6ce116e1674eeae5f0a 2000w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1435&q=90&w=2100&s=a6452b68bcc757a28d1b984073914720 2100w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1503&q=90&w=2200&s=988a3d56a71cf9b5170422f841797158 2200w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1571&q=90&w=2300&s=6ae78cbd26dfa72c406a2fb78a4d3514 2300w, https://solutionseeker2021.imgix.net/images/old/figure2_2021-11-29-185002_wwue.png?auto=compress%2Cformat&crop=focalpoint&cs=srgb&fit=crop&fp-x=0.5&fp-y=0.5&h=1639&q=90&w=2400&s=d6152f72b0460160547c896d2381b03d 2400w)
4
We believe that there are four prevalent challenges that degrade the performance of data-driven VFMs:
- Low data volume
- Low data variety
- The available data have poor quality, they are noisy and biased
- The available data originate from a non-stationary process
In an upcoming series of articles, we will share and discuss recent findings of Solution Seeker’s research lab. These findings are published in an article that you can find in the reference list below. The series begins by questioning the robustness and trustworthiness of the traditional approaches to data-driven VFMs in light of the four challenges listed above. The findings are based on experiments with a large and diverse set of data from 60 petroleum wells located on five petroleum assets. Stay tuned!
- Article 1
- On the challenges of modeling data-driven virtual flow meters
- Article 2
- Inspecting flow rate measurement accuracy with cumulative performance plots
- Article 3
- Inherent challenges of traditional data-driven VFMs
- Article 4
- Calibration plots - how confident is your model about itself?
- Article 5
- Can we trust the data-driven VFMs?
References
- Tareq Aziz AL-Qutami, Rosdiazli Ibrahim, Idris Ismail, and Mohd Azmin Ishak. Development of soft sensor to estimate multiphase flow rates using neural networks and early stopping. In International Journal on Smart Sensing and Intelligent Systems, volume 10, pages 199–222, 2017
- Grimstad, B., Hotvedt, M., Sandnes, A.T., Kolbjørnsen, O., Imsland, L.S.,2021. Bayesian neural networks for virtual flow metering: An empirical study.arXiv:2102.01391