Dr. Marek Kacewicz of the Chevron Energy Technology Company spoke about the development of petroleum systems modeling, from some of the early stages, to the present day, while addressing some of his personal goals and wishes for the future. As he put it, petroleum systems modeling is “an integration of geological disciplines to basically evaluate sedimentary basins.” A majority of the work done by Kacewicz focuses over a broad slate of time scales and regional scenarios. Kacewicz was introduced to the crowd as a long-vetted member of the modeling community, and while he did not start out as a geologist, his background in mathematics helped him focus in on the technical issues facing the development of modeling systems.
While unconventional systems would be discussed in the question and answer portion of the lecture, the primary target of the presentation focused on the sediment compaction and the evolution of porosity and permeability through time in typical sedimentary basins. Those traits are then analyzed alongside factors such as the maturation of the source rocks, hydrocarbon volumes, and pore pressure to give a general and accurate insight into the system. Kacewicz did his best to display a current system while running through the multitude of factors that needed to be considered, highlighting each for the audience.
The fundamental principle of petroleum systems modeling is the interdisciplinary connections that are needed to make a model work. Throughout the presentation Kacewicz was sure to highlight what disciplines were needed to advance a stage of the modeling. “Most [modelers] come from one discipline and learn some others,” stated Kacewicz, and there was no doubt that communication was key to good work. The laundry list of backgrounds listed off echoed of a department list for the school, among the more necessary were seismic analysis, sedimentology, geochemistry, rock mechanics, and petrophysics. As the disciplines have worked together to further modeling, positive benefits have developed. Kacewicz proved this by saying, “From ten years ago to today there has been a change in definition of what we can get.”
Another driving force behind the development of petroleum systems modeling is a progressive increase in computing, from old school machines to the high performance computing and cloud functions of today. Kacewicz started by blissfully recalling some of the lost benefits of old technology. When displaying a slide of an old Cray machine, Kacewicz jocundly revealed, “They were better than laptops because you [could] sit on them in order to think things through.” The separation of man from machine and even machine from office building would be the eventual direction of the history of the modeling machines. Kacewicz then focused on the necessary steps, or layers according to his diagram, that one had to progress through to come to a solution.They are, as Kacewicz put bluntly, “what you have to go through to do petroleum systems modeling.”
The progression of the steps was linear, working through data integration, seismic inversion and interpretation, process description, mesh interpretation and numerical solvers, and finally calibration, uncertainty and risk. These steps are bolstered by high performance computing and security, the latter of which was revealed to be crucial as the art of petroleum systems modeling is put into the realm of cloud computing. Kacewicz went through each facet of the process and discussed the current state of the step, as well as where it is going and how close it is to being there. For data integration it was simple, what was needed is “flawless communication between tools and between disciplines.” Kacewicz was hopeful for the future, but with some disdain he revealed that they were still not at that point.
For data interpretation and seismic inversions, Kacewicz revealed one of the most remarkable bits of technology that was used. “We decided that there are some systems that are so complicated that we want to see them in three dimensions,” announced Kacewicz. What came from that was revealed to be a room where modelers could inspect three dimensional grids and move them around in real space with the use of special goggles and gloves. Towards the future, interpretation still has lacking elements. Kacewicz highlighted the need for better understanding of both salt and sediment evolution as well as the properties of the source rocks.
In terms of his process for interpretation, Kacewicz was frank, saying, “If I don’t get back what is in the well, I go back to the model and I change the transforms.” This has proven to be especially important in terms of faults; if there is an interpreted fault but the data doesn’t match what is modeled, there is something wrong with the fault interpretation. Process description on the other hand was described as a much simpler step. It entails being able to have data resolution at all useful levels in such a way that connections can be made between the involved disciplines.
Although it is a much more technical step, Kacewicz devoted a large amount of time to working through dynamic meshing. The key for the meshing process is to determine if hydrocarbons can find their way to a location in the subsurface. Since part of the modeling involves the release of hydrocarbons from their sources, the pathways that become apparent through the meshing are essential for model based exploration. Just as it has been a key word for Van Tuyl lectures in the past, complexity reared its face in meshing. “If you miss the complexity of the system, it is not good,” explained Kacewicz. In opposition to the complexity, it is also necessary to have fast solutions to evaluate a model as it evolves. Kacewicz finished up with the meshing portion by explaining the necessity for powerful processing in meshing. Where a slow system is easy to model through the meshing, fast systems are laborious and need top of the line machines.
The last major step of petroleum systems design involved a tough question and a more difficult answer. Kacewicz asked, “How do we handle uncertainties?” For the modeling procedure, uncertainties are dealt with by creating maps that indicate the probability of how likely it is that hydrocarbons exist in a given location. The lecture continued with a bit of humor on behalf of Kacewicz as he put up a slide of some extremely complex mathematics. “I would like you to memorize this one,” grinned Kacewicz, “It is not very easy.”
As he had alluded earlier in the lecture, the end was devoted to discussing high powered computing and security. While most of the people in the lecture were geoscientists, there was a sense of amazement as Kacewicz walked through the design of several of his computing systems. Part of his work is devoted to optimizing the software to best utilize the machinery. Hearkening back to the equations from earlier, Kacewicz described the process of designing arrays that could pull in the variables and equations and output them in a timely matter. That lead directly into the discussion on cloud computing as a powerful force in the future. Kacewicz’s hope was that in the future both large and small petroleum shops will be able to use the same material to both side’s advantage. “It is not just a question of storing data, but a question of computing on other’s systems,” Kacewicz said with pride, “Instead of developing things ourselves, we can connect it with others.” Since cloud computing was mentioned as being important to the future, there was also discussion on security. At the beginning of the year Kacewicz had some concern from some of his colleagues over who may be looking at data online. “Who knows who is looking at these things,” Kacewicz reminisced, “Oh wait here is the NSA.”
Before questions, the future of petroleum systems modeling was brought up one last time. Kacewicz wants to see more progression in advanced seismic inversion to the point where it could be done real time, so that drilling results could be seen to build immediate subsurface models.
The focus of the question and answer portion was devoted to linking modern petroleum systems modeling to other disciplines and tasks. The ability for the models to be done on geothermal systems was one of the major points brought up, a concern that Kacewicz dealt with by explaining how they were very much related. The question of the most key uncertainty was also brought up and revealed to be calculating subsurface velocities. Kacewicz displayed some concern over the variables but also stated with confidence his ability to know what was going on, “I have a statistically meaningful sample of projects I have been part of, maybe one thousand… or two… or three.”