Custom-built AI solution for an offshore electrification project
Many oil & gas operators are electrifying their operations to reduce CO2 emissions. Some operators may want to replace an individual gas turbine by solar panels. Others have the operational footprint to consider large-scale integrated?projects.
Such larger-scale projects lead to significantly more complex decision-making due to the sheer number of choices that need to be made.
This article?describes a real project and an innovative custom-built solution that has been developed with Equinor to ensure quality decisions.
Background of the project
Consider 8 offshore platforms, and the option to bring green power from shore?-?about 200km away.
The main project decisions, at this stage, are:
1. Electrify each platform?to what extent??
This choice depends?on?how much CO2 emissions can be avoided over the remaining lifetime of the platform versus the cost to do so. Full electrification will have the greatest impact and the highest cost (large cables).
2. AC or DC power from shore?
The platforms need AC power which can be provided through direct cables from the onshore station. Alternatively, a lower-cost DC cable can be used from the station, but that concept needs an offshore DC:AC conversion hub.
3. Where to place the offshore hub?
The cost of the hub itself is, amongst others, dependent on the water depth in which it needs to be placed. And the hub's position also influences the cables' length?to the platforms.
4. Direct cables to each platform, or "string" multiple platforms on the same cable?
Cable length can be reduced by stringing platforms together on the same cable. But cable cost depends on the required power throughput. And stringing requires more platform modifications.
So why is there value in AI support for this project?
领英推荐
The above decisions result in close to 20,000 different options to consider. Manual workflows break down with these numbers, i.e. cannot investigate the full option space.
Equally important for good decisions: we guarantee a consistent comparison of the options. Manual work carries a risk of suboptimal decisions because scenarios are manually "optimised". By using algorithms we ensure that hub positions and cable routing are actually optimised. This consistency allows for an apple-2-apple comparison of options.
Custom-built AI-powered solution
We supported this project with our AI approach to complex problems. And we crafted a custom-built user interface for the client to evaluate the results and support project decisions.
The project team can evaluate the entire option space through the user interface. Simple user controls and interactive visualisations allow for straightforward, and on-the-fly, comparison of all options.
Our approach also allows users to take a step back and look at?key decisions more?holistically.
For example, the AC versus DC from shore decision can be evaluated across the full option space. DC?from shore?is?often?cheaper?because that concept?needs?fewer cables with higher capacity instead of several AC cables.?But this is not the case in all scenarios. More insight can be derived from evaluating the full option space.?See the below cross plot of Cost/CO2 abatement vs CO2 abatement - with light grey data point being the DC from shore options and dark grey the AC from shore. This plot immediately shows?in what specific scenarios?the DC from shore concept becomes more?capital efficient. It can also be used to show in what scenarios there is little to no cost difference between both concepts, hence where other considerations (such as execution risk) may be the differentiator.
Users can also immediately see how the optimum hub position varies across all the relevant options, see the red points on the map in the above visual. This may help the team acquire additional sea bed surveys at particular locations early in the project.
Closing
This article describes the complexities of an offshore electrification project. And we have described how a custom-built AI-powered solution can support decisions.
Most large-scale infrastructure projects have similarly huge and complex option spaces. Supporting such projects with manual concept engineering workflows is not good enough. There may not be any existing software that captures all the relevant detail. But that is really no excuse to accept suboptimal decisions. Custom-built solutions can be made in less than 2 months and are often very cost-effective when considering the risk of the project.
We are grateful for our strong collaboration with Equinor, and the opportunity to develop innovative powerful solutions together.