French software house Kappa has launched Kappa-Automate (K-A), a containerized version of its well test suite for the modern IT world of microservices-based workflow automation. K-A adopts a similar software development pattern to OSDU, the Open Software Data Universe, but Kappa’s development preceded OSDU by a couple of years and the toolset is now deployment-ready.
As Kappa CEO Olivier Houzé told Oil IT Journal, ‘The Automation project started with a KAPPA consortium in 2016, after requests from our major clients to move beyond permanent downhole gauge preprocessing and integrate some of our software functionality with automated workflows. OSDU began in 2018 and deals with a much bigger picture, but the objectives were similar. It was natural for us to join OSDU when it started gaining credibility’.
‘Automate is exactly what you describe: a self-contained automation architecture which deals with the workflows related to our software portfolio. It can have its own life and be installed and operated on its own. But we expect that it will be more like a satellite of a larger OSDU compliant structure. In the latter case, the Automate structure will be transparent and it will be as if third parties were directly using our microservices. Under the hood, it will be a bit more complicated!’
‘The OSDU R1 and R2 releases were not directly related to our segment but things are changing with R3 and the integration of Schlumberger’s Delfi platform. R3 will cover the data that interests us, so we anticipate that we are going to integrate the OSDU process during the development of R3 and that Automate should be OSDU R3 compliant when released in the not-too-distant future’.
In an online video presentation, Houzé traced the history of well testing in the context of an industry prone to periodic downturns. Automating analyses began as early as 2000 when the deployment of permanent measurement led to information overload. Kappa’s software evolved to help engineers by identifying events of interest and sending data on for analysis. The 2014/15 downturn and the ongoing big crew change has led to a ‘perfect storm’ of human and technology disruption, compounded by a demographic gap that resulted from the no hiring policies of earlier crises. Engineering expertise has gone from many operating companies.
On the positive side, technology, especially artificial intelligence, holds promise, even if there is some controversy around AI’s potential and limits. AI that identifies a squirrel as a sea lion (with ‘99% confidence’) is not going to replace the engineer.
The technology under the hood (as in OSDU) represents a shift from the monolithic apps of the past to ‘small pieces of software’, a.k.a. ‘microservices’. These can be installed anywhere, on servers on-site or in cloud. K-A provides microservices for pressure and rate transient analyses and more and is said to present a great opportunity for interface with other vendors’ solutions via the upcoming OSDU R3 release. The first commercial release of K-A is scheduled for Q4 2021.
On the application front, Kappa is cautious about the move to the cloud/web/HTML5 paradigm. It took some five years to see the benefits from the previous migration from MFC to .NET. It remains to be seen if HTML5 software can be as rich and user friendly as Windows. The first Kappa tool to benefit from the Automate interface is Carbone (PVT analysis), part of the ‘next generation’ G6 release, now deployable from the cloud.
In the video, Houzé commented that Saphir (Kappa’s well test flagship) is ‘a bit big for microservices’ which echoed earlier considerations that we have encountered for instance in our exchange with EnergySys’ Peter Black. We pressed Kappa on what exactly constitutes microservices and asked for some examples of how they might be deployed. Kappa’s Olivier Allain came back with the following.
One example is the ‘incremental PTA*’ workflow. Traditionally, an engineer takes high frequency raw downhole pressure measurement and production data and executes a build-up analysis and outputs a PTA analysis file. An automated workflow replaces this manual analysis as follows. A filtering service preprocesses the raw pressure data which is fed into a processing service to identify shut-in periods. For each shut-in, a PTA service is called to load data and diagnostic plots and rerun and improve a pre-selected model. The results are fed into time lapse and/or map displays or perhaps on into another workflow.
This may just be a matter of semantics but are Allain’s above ‘services’ microservices? Or more to the point, will they ever be used outside of Kappa’s automated workflows? There must be a lot of folks in the OSDU community struggling with similar questions. KAPPA expects answers to these questions in a 2020-2021 K-A field test, where several operators will be trialling K-A in their own workflows. The Kappa Automation Consortium (KAC) currently has 13 members including BP, Chevron, ENI, Equinor, ExxonMobil, Marathon, Shell, Aramco and Total.
* Pressure transient analysis.
© Oil IT Journal - all rights reserved.