The Artificial Lift R&D Consortium’s 2017 Gas Well Deliquification workshop took place earlier this year in Denver. Theta Oilfield Services’ Terry Treiberg traced the ‘pendulum’ of artificial lift software beginning with the majors’ in-house developments of the 1980s and 1990s. The early 2000s saw the emergence of platforms from Wonderware, Case Services, Cygnet and XSPOC. Followed by data warehousing historians, notably OSIsoft’s PI System, along with visualization tools.
Treiberg noted the ‘lure and disappointment’ of analytics with a proliferation of machine learning, predictive analytics and IoT offerings with exaggerated claims for ‘deep and predictive insights.’ The problem with these approaches is statistical models alone are ‘average at best, and useless in many cases.’ Domain knowledge is needed and data must be correlated, interpreted and understood in context. Statistical methods can help but they are at their best when combined with proven engineering algorithms. Treiberg illustrated such domain smarts with a variety of downhole pump dynamometer card signatures illustrating different situations. Artificial intelligence has a role to play here in a hybrid approach that combines scada data flows and analytical engines with ‘customer-configured, open, extensible and integrated tools.’
Bill Elmer (Encline Artificial Lift Technologies) asked ‘what does the internet of things mean to the oil and gas industry?’ The IoT’s goal is to turn everyday production equipment into a smart device that allows a lease operator to remotely ‘check-in’ to equipment to monitor indicators of interest from a webpage or app. IoT reality is closer to the E&P industry than you might think. Devices communicate using Modbus TCP and conventional scada programs can pull or push data and communicating using remote procedure calls. This provides the ability to perform engineering calculations at the wellsite and to create equipment KPIs, alarms or shutdowns based on calculated indicators. But what of security? Elmer warns against listening to the sirens of the IT industry. ‘You don’t need a massive cloud-based system.’ Operators just need to maintain private internet networks behind firewalls that are ‘never in contact with the cloud!’ ‘Keep your data private and never share outside of the internal network!’
The ALRDC meet also heard from Rosmer Brito on a proposal to create a production laboratory at the Colorado School of Mines. The School is looking for relatively small donations to create test labs covering downhole and surface facilities. More on the CSM proposal here.
More presentations from the ALRDC.
© Oil IT Journal - all rights reserved.