Matlab real time framework

Wintershall and TNO Science and Industry eschew vendor-based solutions, preferring a Matlab-based, object-oriented software framework for production monitoring and perhaps, control.

Speaking at the SPE Intelligent Energy conference in Utrecht, NL last month (more on page 6), Hans Reijn provided an in-depth look at Wintershall’s flagship Integrated Production Management System (IPMS), a key component of its €11.5 million Den Helder, NL-located Remote Controlled Operations (RCO) center, claimed as one of the most modern platform control rooms in the world. Southern North Sea assets are hooked in to the RCO via microwave radio.

The IPMS, co-developed by Wintershall and TNO, was designed to address production from increasingly complex, small reservoirs in the face of a liberalized EU gas market. At the heart of the IPMS is a model-based advisor that uses real-time data to align production with market conditions. The advisor provides event detection, condition monitoring and helps operators ‘make informed decisions that add value to operations.’

The Advisor is a classic case of model-based optimization, where real data is continuously compared with the output from mathematical models of the component physical subsystems. A divergence between the two indicates some abnormal condition—which can usually be identified by comparison with a library of known behaviors. The system combines TNO’s experience of model-based monitoring and optimization using a Matlab-based system architecture with Wintershall’s asset models and real-time infrastructure. Real time production data streams to the OSIsoft PI System historian and is compared with models of critical asset components including reservoir, wells, flowlines and surface equipment.

The system covers the entire asset, from reservoirs to top-sides. Combining the large number of individual components in software requires an object-oriented framework and a configurable architecture. The system also handles the frequent changes in asset configuration. Matlab and third party toolboxes provide functions for statistics, filtering, modeling, and data visualization. TNO has extended these with an in-house developed toolbox for reservoir, well and equipment modeling.

Matlab is now the basis of the real-time monitoring application and connects directly to PI System. A library of Matlab routines targeting production modeling and forecasting has been developed. These can be called from Matlab itself, or from other industry models such as GAP or MBAL. Reservoir and wellbore models developed in Excel can be incorporated into the real time environment. Reijn’s paper (SPE 128300*) describes in detail various component calculations including material balance, flow rates, choke settings, inflow performance and more. Model use cases include salt precipitation mitigation by water wash using an IMPES simulator.


Click here to comment on this article

Click here to view this article in context on a desktop

© Oil IT Journal - all rights reserved.