How GPM made its way into the Petrel platform. A technical journey… or is it?
Stardate 2010
SIS VP of technology (Meyer B) asks myself and Rod Laver (another getting-grey-hair-guy at the time) to go find some geoscience software that will make a difference to our marketplace
We travel to Stavanger to see what they have in the Stavanger research centre—we meet the usual SSR/charisma-wizards—masquerading-as-Petrelites. They took us over all the ‘usual-suspects’ technology e.g. around automatic horizon extraction—something they had shown me 15 years earlier and sometimes it seemed we were in a kind of ground-hog-day, apart from something they had to do with forward stratigraphic modelling.
Around this time, Per Salmonsen (senior charisma and geoframe, wizard who also looked after all development of the Petrel E&P software platform for a number of years) started looking at a thing he called GPM – geologic process modeller. It was based on some code we had in research for many years prior, developed by Dan Tetzlaff – who wrote the first 3D forward stratigraphic simulator back in the 80s, as part of his PhD at Stanford, and thankfully was still working for us. The code we had was kind of a re-incarnation of that. We talked about it quite a bit and thought that perhaps this kind of modelling, while previously considered too hard to parameterize, too difficult to find the right input data, too slow, and too hard to tie to wells, might be coming into its time soon.
We certainly knew that many of the majors and larger NOCs with research groups were looking at these workflows and trying to see if they could be applied to improve reservoir understanding and predictability beyond the current state of the art, which was largely based on geostatistics based models – for many kinds of complex reservoir architectures – such as turbidites. This is where our research to-date had been concentrating on, and it was a very popular theme with our operators.
However, we soon began to realize that there were many groups looking at carbonates, where depositional sedimentary concepts are only part of the story. Furthermore, none of the major competition had any expertise or significant offerings in this arena, and we of course could also offer integration with other parts of the Petrel platform. So while we went ahead with pushing a general forward stratigraphic modelling tool, we thought there was a good business case to have a specific focus on the characterization of carbonate fields—where lots of the world’s oil is tied up, but the reservoir architectures are often poorly understood.
However, there was lots of competition from other internal projects vying for engineering funding, so we were not given a blank check, but at least over the years, a series of enlightened management folks gave us the green light to move ahead.
Circa 2011
A proof-of-concept plugin was developed with some help from the Tyumen tech center.
2012
We started looking around for likely partners and sponsors. Per and I travelled to the Middle East and the USA and presented to a number of the major operators in those places. We thought we had struck gold in the Middle East, when one of the NOCs had a research group that was working on the same thing and expressed great interest. After many long flights and meetings and discussions of what we could and would do, we were at the point of a deal where they would actually invest in 50% of all the development costs.
In parallel, we worked with the Aberdeen based ITF group—a consortium building org, funded by many of the operators, to bring new solutions to the market for various themes—forming consortiums between operators and research teams. Various companies from that avenue dipped their toes in the water but pulled out almost as fast. A common perception is that scalable-and-usable-by-asset-geologists solutions are readily produced by internal R&D groups, if you fund them enough or long enough. I met with many of these groups at this time through this channel, but most struggled to work with other teams outside of their own organization. Collaboration is hard.
However, from this exercise we did manage to catch the interest of Wintershall, who eventually became one of the founding members of our advanced GPM consortium, which we called the GPM JIP (joint industry project). This is 100% due to the drive of Dr Peter Suess at Wintershall, who has held various senior geology and management positions there—and continues to help drive the JIP today, and the discipline in general, as well as being an adjunct professor at the University of Tübingen.
In one of our visits in Houston to all the well-known major operators, as we were presenting what we were trying to do with GPM, one of the operator’s geologists piped up and said something along the lines of ‘this is all very well, but if you guys really want to know how to do this, you need to read this book’ and promptly slammed down a book on the table. Fortunately, Dan Tetzlaff being based in Houston at the time, was with Per and I in the meeting and said simply ‘I wrote that book’. The geologist looked at him, looked at the book and then said, ‘Can you autograph it for me!’.
Early 2015
Things all seemed to be heading in the right direction, up until early 2015, when the wheels fell off of the entire industry. Our Middle East research group that was set to fund us up to 50%, lost their budget and Schlumberger, like all of our clients had to trim our costs. Certainly, interest from clients in sponsoring new workflows seemed to be suddenly much harder.
But then came along Dr Olinto from Petrobras. From a contact with one of our enterprising geoscience business owners in Brazil, we began a long and fruitful partnership with Petrobras’s central geology modelling team in using and developing GPM to assist them in the characterization of their subsalt carbonate fields. That partnership is still running today. This partnership owes a lot to a little serendipity and some very good work.
It turns out that Dan Tetzlaff had, up until recently, been used as a Petrel modelling software engineer, concentrating on facies modelling, kind of biding his time and waiting for the world to wake up to the power of GPM. Dan was originally from Argentina, but even so, could communicate easily with his northern neighbors. Dan is also blessed with a kind of easy-going paternal nature as well as an endearing self-deprecating approach to his own genius. All this made him perfect as a trusted advisor for Petrobras, which is still the case.
Thanks to the funding from the Petrobras project and the ongoing support of our VP of portfolio at the time (Trygve Randen) the project survived the downturn.
2016-ish
Not too long after, I happened to be giving a presentation at a Schlumberger technology day in Vienna and I showed some of the models that we had been developing with GPM. I was approached afterwards by the chief reservoir engineer of OMV who said the coupling of this kind of simulation with dynamic reservoir simulation was exactly what he was looking for, as to understand the behavior of their very complex field—a serious stacked deltaic sequences—with every new well (in addition to the hundreds they already have) being a geological surprise. After a short trial of the software, they signed up for the GPM JIP and we had another member. More on this one later.
In the meantime, Per had been working on multiple fronts non-stop for all those years to get the product ready for a commercial release but also to raise awareness of the value & to train our people in what this thing was for & how to use it. By hook & by crook he managed to cajole young Schlumberger geologists to use it as part of their ‘GFE’ – and kind of coming-of-age project in Schlumberger – in order to create example models – and various young geos from the UK, Norway & especially our Aachen tech centers started building some very cool models, based on papers, outcrops, public data sets etc, which have become invaluable in being able to show the power of GPM.
He also was snaffling people from various technology centers to donate their time to develop bits & pieces here & there, test the code etc etc. It reminded me of the fable of the guy who tells you he can make a fantastic soup just from nail (as long as you can borrow potatoes, onions, carrots, herbs, salts etc etc). Pretty soon we had a commercial grade piece of software, tightly integrated with Petrel.
But that was not enough for Per. Having vast experience of how to build scalable, complex software systems, he could see that GPM could easily get bogged down if he chose a poor software architecture to build it upon, so he set about designing a whole new framework system for GPM, in partnership with Jan Tveiten, a long standing trusted colleague, and software engineer – one of our ‘gun programmers’ if you like.
This was also not enough for Per. He knew we did not know enough about this subject for us to realistically create a good product. Per came from a geophysical technology background – picking up a few geology concepts along the way. However from 2011 onwards, Per starting reading more and more papers & books on the subject of forward stratigraphic modelling, carbonates & geology. Over the intervening years he has kind of given himself a self-taught PhD (or more) in this field and is now key to the geologic direction we find ourselves pursuing with this project.
2017
We are approaching our first commercial release of GPM in Petrel, We have several companies doing service projects with the pre-commercial version of GPM, including our JIP clients, who had an ‘advanced’ version of GPM.
Per is merciless in chasing management to invest in training of people – perhaps because they were tired of hearing us pushing them or because they really thought it was a good idea, they allow us to run live training classes on GPM, with HQ paying for the travel & living of the attendees who came from around the globe.
While my most recent decade or so at Schlumberger has seen me report to veritable procession of managers – on average a new one every 18 months (perhaps that says more about me than them ☹ ), all have been encouraging & supportive of our quest to get this technology to the market. In particular, the VP of Portfolio at this time Mark D, embraced the vision for GPM & gave us the support we need to get the product commercial and deployed at least in the minds of our geoscientists and account managers around the world. This was not just via the special training classes but with airtime & workshops at our internal sales & marketing communications events ‘M2S’ as well as our annual internal global product training event ‘PTU’.
All this helped in ‘spreading the word’ and not long after we had people popping up all over the place asking for trials.
Enter Sergio Courtade
Sergio actually came into the GPM team a year or so earlier as the product champion. I first knew of Sergio as a guy who used to run the Petrel commercialization team. A kind of slow talking Argentine dude but he knew ‘a bit’. Around 2015/16 Sergio worked for me as an expert in quantitative interpretation & we would deploy him to the far corners of the earth to teach QI to various clients. I lost sight of him for a little while then he popped up under Per as the GPM product champion - and it wasn’t until then that I discovered he was a former professor of sedimentology – so with that skill set & already being a Petrel wizard he seemed like a good fit. It seems I underestimated him. He took like a duck to water with GPM. Before the OMV evaluation started in mid 2016, he asked them for any paper or analog they had on their key field. They sent a paper referencing a present day river-delta system in Queensland Australia as an analog. 2 weeks later he gave them a live presentation over Skype, having modelled that Australian delta with GPM, explained in great detail how he modelled the various elements of the deltaic architecture and explore different scenarios. I believe that presentation was what moved us rapidly from an interest in exploring GPM to a full blown collaboration, that continues to this day.
We release GPM with Petrel 2017.2. Sergio gets rather busy and starts doing evaluations with clients all over the pace – but especially in the far east. Using his geophysical background, he also introduces the workflow of forward seismic modelling the forward geologic model to compare with existing seismic. With this he explains what various features in the seismic may be stemming from in the geology &/or what geologic features may exist but can not be seen in the seismic – this workflow creates quite a buzz!
While in my home country of Australia & while presenting to a well known operator in Perth, two interesting comments were made to me.
-
Firstly from the chief reservoir engineer – something like ‘this is great – I am trying to show my engineers that not exactly tying all the wells is a lot less important than landing a reliable geological concept that will better tie new wells’ – Like our friends from OMV his observation was ‘ every new well we drill is a history matching surprise’.
-
Secondly from the chief geologist, along the lines of ‘please don’t show this to my team – this is like pornography for geologists and they will want to do nothing else !”
2018
As part of Per’s relentless interrogation of the literature he soon realizes that while many universities a re performing research in the manner of GPM, there are a few that stand out with a proper track record of knowing what they are doing – and only a very few are doing that do that in his passion – modelling carbonates. One in particular was a group at the University of Bristol, led by Prof Fiona Whitaker. Around 2017/17, with them being within easy striking distance of London (where my office was) we organised a couple of meetings with Fiona to discuss collaboration. Nothing much came from that, except some goodwill.
In parallel, another well known middle east based NOC has had a ‘research competition’ running for many years, looking for collaboration with interested parties on various themes - carbonate diagenesis being one of them. I had personally been involved in several attempts in the past to raise funding for various projects via this channel. On this occasion we decide to invite Fiona to team up with us to write a submission to further develop GPM’s capabilities in diagenesis and apply it to the operators fields. This we did in early 2018. I remain completely embarrassed when I compare the quality of our final submission to the puerile, unsuccessful attempts I had participated in previously. Needless to say, we succeeded and that not only open up a new relationship with the Uni of Bristol, but also opened other doors for us with many of the world experts in this field having previously studied with Fiona, suddenly giving us the time of day.
2019
After a very successful GPM JIP meeting in Montpellier, including several of the world’s leading operators and experts, one of the world’s VERY LARGEST operators also joins the JIP, after a combination of excellent support from our other members, Fiona and another brilliant evaluation-training over the wire from Sergio.
We learn a lot from our JIP partners about the value of GPM – especially in the meeting where we have a lot of time to talk on field trips, over dinner – and of course in the bar. One key theme that has emerged, in it has tremendous value in just teaching people how to think geologically, and evolving good, physically plausible geologic concepts. Making very sophisticated, often visually striking models is of course very compelling- and the re-use of the models in other Petrel workflows is invaluable – but the value of landing a good geologic concept seems to be king.
As a consequence we have advanced plans to release a very fast, very simple 2D version of GPM, which will make the method accessible to every geologist – with a quick learning curve and instant results. This ‘GPM2D’ is currently in ‘field trials’ and we will release to the market later this year, and the journey will continue, hopefully with a bright future, giving geologists the opportunity to learn more about their chosen vocation, make better decisions, and have fun!
End of Chapter One
Author information: An Australian and ‘lapsed geophysicist’ based in London, has been working in the E&P industry for more years than he cares to remember. Alex spends a lot of his time connecting people together to build projects that push the boundaries and improve the science that we are all using—a kind of human glue-gun.
