You’ve heard of the semantic web I expect—Berners-Lee’s pipe dream of a data driven web—enabling machine to machine communications and pinpoint search. I use the word pipe dream advisedly because the semantic web at large has been as big a flop to date as the original world wide web was a success. The main reason for this is that while folks are happy to cobble together some forgiving code like <h1>My Big Headline</h1> etc., the instructions for deploying the semantic web are hard to find, generally incomprehensible and requiring of an effort well beyond the average web developer—who is busy trying to code for three different browsers, none of which support the semantic technology anyhow.
But, the argument goes, this is not so in the enterprise, where browsers can be standardized and extended and where code can be written by a machine—hiding the ugliness. Moreover, at the enterprise level, it should be possible to mandate, although industry has not been very good at this to date, compliance with a standard representation that facilitates data sharing, without resorting to a ‘standard’ (i.e. proprietary) solution from a single vendor.
It would be even better if this happened at the national level—so that a region’s companies and suppliers were all singing to the same semantic hymn sheet. And this is exactly what might happen in Norway, if the POSC/Caesar association (PCA) manages to pull it off (see our report from the PCA ‘Semantic Days’ conference on page 6). Semantic web technology has been adopted for deployment of the ISO 15926 standard for the description of plant and process hardware including offshore structures. Some now see the approach expanding to embrace more upstream protocols such as WitsML and ProdML.
ISO 15926 is also getting traction in the US through the Fiatech standards body (see page 3). Here BP, Chevron and Petrobras report on work that involved the standard, albeit not all at the enterprise scale. As usual in the standards game, some ‘deployments’ may be restricted to the use of a ‘compliant’ application. Nonetheless, it is pretty significant that a US standards body is getting in on the act.
In fact it is fair to say that the PCA/Fiatech activity, if it does develop as planned, will be one of the first truly significant semantic web deployments in any vertical. Which is both worthy of praise and also a bit scary!
~ ~ ~
Last month we reported from Phil Crouse’s PNEC data integration conference in Houston and the SPE Intelligent Energy show in Amsterdam. I was struck by the difference in tone of these tradeshows. As the Intelligent Energy show is about applying digital technologies to the oilfield and PNEC is about managing the data that comes, inter alia, from the oilfield, you would think that these shows reflected different sides of the same coin.
How is it then that Intelligent Energy is a razzmatazz celebration of ‘cool stuff’ and PNEC is, well, a bit tedious? How come Intelligent Energy (IE) portrays a world of successful deployment of projects of mind boggling complexity and reach, while the poor old PNEC crowd is still complaining about under resourced projects, poor quality data from vendors and so on? And most importantly, how can all these collaboration rooms, fields of the future, model-based decision support systems claim to work properly if the data that they depend on is in the parlous state that PNEC would suggest?
Part of the answer lies in the sneaky ‘inter alia’ of my first paragraph. PNEC, for historical reasons, is mostly about geological and geophysical data management. This is because the G&G community began looking at data management in a serious way a decade or so ago. The petroleum engineering community has only really been into data management for a couple of years.
But that’s not quite right either. The digital oilfield is lucky compared with the upstream in that it had a large installed base of process control instrumentation and systems. The engineers didn’t really need to manage data, it was already being done for them in the historian with a large installed base in process control.
But that’s not the whole story either. In his keynote at IE, Schlumberger CEO Andrew Gould cautioned that although the concept of real time optimization has been around for a decade or so, along with the expectation that recovery rates would be boosted, this has not really happened. And this is despite a history of digital oilfields going back to Exxon’s Computerized Production Control in 1967!
So with such a digital history, why all the histrionics? There is a degree of commercialization at IE which is less evident at PNEC—something of a paradox when you think that IE emanates from the Society of Petroleum Engineers. The commercial razzmatazz means that even the more technical presentations get sucked into an excess of ‘wow’ factor. PNEC talks in general tell it like it is.
How is it?
And how is it you may ask? Well you can‘t do the digital oilfield without sorting the data issue. So one has to imagine that much of the singing and dancing of IE concerns somewhat early stage development. It’s much easier to develop a skunk works solution that flies for the time of a demo, than to assure enterprise wide success for a relatively modest project. Another issue impacting the upscale conferences these days is the skills shortage. If you are trying to recruit you need to make it sound better than it is. If you are trying for more funds, make it sound worse. Perhaps PNEC should be considered as a user group and Intelligent Energy as a beauty contest!
© Oil IT Journal - all rights reserved.