|OPPORTUNITIES AND OBSTACLES FOR INTEGRATION OF SUB-SURFACE TECHNOLOGY AND FACILITIES DESIGN|
|Field Developments and Case Studies|
1Petroleum Development Consultants
The potential for integration of sub-surface technology and facilities design has been recognised for many years although the software tools needed for this were not available until quite recently. For reference, the value of integration, and in particular between static geological models and reservoir models, has been recognised in the last ten to fifteen years.
Due to the lack of available tools to integrate sub-surface technology and facilities design, work in this area has in the past generally been carried out by the oil companies using their own proprietary software. One of earliest attempts at this involved the connection of a two-phase gas reservoir simulator with a surface network simulator by Amoco to mange the production of a number of Southern North Sea gas fields(1).
In the last ten years commercial software providers have developed the integration software to link well known commercial software products using the Parallel Virtual Machine interface. The interface solution offered by Schlumberger called Open Eclipse allows the linkage between Petrel, Eclipse and Hysys (process simulator). Another software provider Petroleum Experts allows the linkage using their Resolve programme between Eclipse, Prosper (well performance software), Gap (surface network) and Hysys.
For the non-engineers these developments might appear not to be relevant to their work. However future production of more difficult reservoirs in more remote locations will mean that surface constraints will increasingly impact the appraisal and development of new discoveries. A simple example may assist in understanding this.
Three of the key determinants of gas field commerciality are the size of the reserve, its location and gas quality. Sales gas specifications usually require a carbon dioxide content of less than 2% by volume. In the example a 300 bcf field with 2.5% carbon dioxide content could be considered uneconomic. In a remote location a 150 bcf field with 0.1% carbon dioxide could be uneconomic. However combining the fields would result in total reserves of 450 bcf with 1.7% carbon dioxide content which could be economic.
Another important area that exploration geologists should understand in terms of the impact of surface facilities is in the development of gas condensate fields. Whilst generally there is a reasonable understanding of what recoverable reserves are likely to be for oil and gas discoveries this is not the case for gas condensate fields. In particular the reserves of condensate and gas (to a lesser degree) vary significantly dependent on whether the field is produced by depletion or whether recycling is to be used. Condensate recoverable reserves in a rich gas condensate reservoir could vary between 20% to 60% of the condensate in place.
ENI in a paper given at the EAGE Annual Conference in Rome in June 2008(1) suggest that the following issues can be addressed by the integration of reservoir models and surface facility models:
A major obstacle in developing this type of integration is the level of detail in the various underlying models. This is a very common problem in often needing to simplify Petrel models in order to run the geological model in Eclipse. There are similar problems between Eclipse and Hysys as the latter uses more hydrocarbon components than Eclipse. However this has been solved in most projects so that reducing the level of detail does not materially affect the accuracy of the results.
An obstacle that is more difficult to overcome is the gap between the reservoir engineer and the facility engineer. This is surprising given that they both cover the topic of fluid flow which is a key component of both reservoir engineering and process engineering. This could be contrasted with the high degree of cooperation between reservoir engineers and other subsurface technologists such geophysicists and geologists. It maybe the key issue here is uncertainty and this is a factor that has to be taken into consideration by all subsurface geoscientists. Indeed it may be an important factor in the subsurface/surface interface that there is a transfer of subsurface uncertainty to the facility engineer in an explicit way.
A further obstacle is the training that subsurface professionals and facilities engineers receive at university and the lack of integration at this level. Integration of subsurface studies (and particularly projects) is now widespread in most universities that offer oil industry related courses. However there is very limited integration between reservoir engineers and facility engineers (generally chemical engineers in the academic world). Again this is surprising especially considering particularly in the UK what a high proportion of chemical engineering graduates are employed by the oil industry.
The use of an integrated subsurface/surface approach appears to show identifiable gains in recoverable reserves and production. ENI in the paper mentioned above note that the use of an integrated approach resulted in an increase of 6.5% of oil that could produced from an offshore oilfield by better allocation of gas lift rates than was achieved by standalone reservoir models.
Halliburton in a paper (2) given in 2008 about how integrated studies help asset teams make optimum field development plans give the example of a giant onshore field with over 1,000 wells. An integrated optimisation model was developed that included production wells, injection wells, the oil gathering system, field manifold stations, central processing facilities and the injection and gas lift distribution systems. The gain indicated in the paper was a 14.2% increase in oil production without significant capital expenditure.
Petrobras and Halliburton in another paper (3) given in 2008 looked at optimisation of surface networks and platform location using a next generation reservoir simulator (NGRS) and an integrated asset optimiser. The NGRS allows data generated in the surface network to be incorporated in the reservoir simulator. The example application involves a synthetic field similar to a North Sea field in terms of its structure, faults, permeability, porosity and subsea conditions. The field was planned with nine producers and three injectors in 500 metres of water using a floating production facility. The objective was to find the optimum platform location. The best case of the 200 simulations run achieved a NPV 4.6 times larger than that of the original location suggested.
I would like to acknowledge the participants at the SPE Applied Technology Workshops on the Reservoir/Facilities interface held in London in 2009 and Aberdeen in 2010 who have contributed to increasing my understanding of the issues involved.