Game Theory Models
In 1950 Merill Flood and Melvin Dresher of RAND (a policy think-tank) concocted the elements of what we now know as The Prisoner’s Dilemma. This is a game theory construct whereby two criminal conspirators are questioned by the police in separate rooms. Each prisoner is rewarded or punished according to how he and his accomplice respond as shown in the matrix below. Although ‘deny, deny’ is the best overall scenario (from the prisoners’ view) there is always the temptation for each prisoner to confess in order to better his own position. The point is that if both prisoners seek to benefit themselves independently they will each be worse off than they would if they cooperated successfully. Trust from both parties is therefore essential for an efficient outcome to result, leading to ‘deny, deny’ scenarios where a total loss of only 2 years arises.
This game demonstrates why a collaborative approach to problems is optimal, and its lessons are widely applicable in political and commercial settings. In light of recommendations from the Wood Review (Feb 2014) about the need for greater collaboration, it may pay for UK North Sea operators to cooperate, particularly with respect to the hefty decommissioning effort that looms ahead. One area where the need to collaborate comes into effect is when there are conflicting field interests in a hub scenario. In 2010 Palantir consultants Willigers, Prendergast, and Muslumov published an SPE paper titled ‘North Sea Dominos: The Economic Dependencies of Infrastructure Assets and Their User-fields’. That paper highlighted the problem of fields, who share hub infrastructure, closing when they can no longer profit as an isolated entity under a simple production-based cost-share arrangement.
The operator’s dilemma
Consider a situation where Field 1 and Field 2 share hub infrastructure that is operated by Field 1. The hub—and consequently, both fields—may close as soon as Field 1 becomes unprofitable based on a hub cost-share arrangement. This is obviously not a problem if Field 2 became unprofitable prior to that time. However, if Field 2 stands to make considerable profits contingent on Field 1 keeping its infrastructure open, we might think it lies in both fields’ interests for Field 2 to pay Field 1 out of its profits earned, in return for Field 1 staying open. It is also in the interest of the UK government to extract as much oil and taxes as possible—so how can this be achieved?
Let’s look at some numbers. The table below represents a fictional hub where Field 1 operates the hub infrastructure and Field 2 is a user of that infrastructure. Field 2 could represent multiple fields if we consider it as an aggregation of all other fields using the hub infrastructure. If we assume an oil price of $90 per barrel along with annual infrastructure opex of $20m fixed costs plus variable costs of $8 per barrel, Field 1 will become unprofitable in 2024 so its COP (Cessation of Production) year is 2023. However in a hypothetical cost-share setting Field 2 would remain profitable until 2028—so the hub is foregoing five years of Field 2’s potential profits.
If it were contractually possible, it would be in the interests of both Field 2 and Field 1, for Field 2 to offer Field 1 some of its profit in exchange for the ability to continue producing until its COP date of 2028. After crunching the numbers, the chart below summarises this scenario—showing that total operating cash flow is at a maximum for Field 1 in 2023 (the effective shut-down date for the hub), while the same would occur in 2028 for Field 2; and when both fields are considered as a single contractual entity, the COP for the hub is 2027 (See EndNote 1). We need some contractual mechanism that will allow Field 2 to re-channel some of its profits back to Field 1 so that the Hub COP of 2027 can be realised, resulting in additional profit equivalent to the triangular area enclosed by points A, B, and C on the chart.
Possible solutions involve working out a schedule for cost reimbursement from Field 2 to Field 1 over the life of the assets. Field 1 would need to demand a minimum profit threshold in order to keep the pipes flowing and the hub running. But such a floating rate arrangement would be difficult to draw up and regulate contractually. An alternative technique is to return to a tariff mechanism—but the key here is to optimise the tariff with Field COP dates acting as a fluid input to the tariff setting process.
In search of a ‘deny, deny’ tariff
One possible method for ensuring hub COP is achieved is to find a tariff that will result in both fields having a positive operating cash flow up until the COP of 2027. This more economic abandonment can be achieved by allowing Field 1 to bear 100% of the infrastructure cost and then charging Field 2 a compensatory tariff. The chart below shows how the NPV of both fields and the hub vary depending on the tariff Field 1 levies on Field 2. The calculations account for Field COP changes as the tariff places extra costs and profits on each field.
The chart below shows that by levying a tariff of $65 per barrel (See EndNote 2) in this scenario the combined NPV for both fields on a tariff contractual basis (Hub COP-Linked) is exactly equal to the NPV for the hub considered as a single contractual entity. However, note the inverse relationship between the NPV’s of Field 1 and Field 2 as tariff increases. The shaded area represents all scenarios where the combined NPV on a tariff basis is within $3MM of what it would be if the hub were considered as a single entity. This may be considered as a ‘negotiating space’ where little or no gross loss of production or profits occurs. Increasing the tariff through this region will simply increase the value that is transferred from Field 2 to Field 1.
Other points to note from the above chart are as follows:
- Kinks in the trend lines occur due to Field COP shifts as the tariff increases
- Field 1’s optimal tariff (or ‘confess, deny’ tariff, to hark back to the metaphor) of $70 per barrel is actually suboptimal for the hub. The drop in overall value occurs due to Field 2’s COP shifting three years earlier as a result of the tariff increase from $65à$70.
- Field 2’s maximum achievable NPV ($216MM) is about the same as what it was obtaining under a cost-share arrangement ($210MM), even with the premature hub COP date under that arrangement! And this is only achievable where the hub NPV is about $50MM lower than it would be if operated as a single entity.
We are nearly there. We have a contractual mechanism that can be used to achieve maximum NPV for the hub—but Field 2 will still opt for cost-share since that alternative provides it with more value. This is because Field 1 must impose a hefty fixed tariff from day 1 in order to come out profitable in the later years through to 2028. To remedy this a sliding scale tariff could be implemented. What may be more straightforward, however, is a ‘costshare to tariff’ arrangement, whereby the tariff only kicks in after Field 1 starts to become unprofitable (i.e. it’s initial COP). The chart below implements this solution.
Like in the previous chart, the shaded area represents tariffs that result in a combined NPV (i.e. the sum of NPV’s for both fields) that is within $3MM of the hub NPV when the hub is considered as a single entity. The difference this time is that throughout the shaded range (comprising tariffs from $50 to $65) both Field1 and Field2 are better off than they were under the cost share scenario (represented by the horizontal segments of both lines (See EndNote 3) in the chart below).
Collaboration is the watchword
The solution above works—however the case presented here is relatively simple, and modelling outcomes for multiple fields with working interest arrangements is tricky. On the positive side the mechanics explained here are not too complex and simplicity is an asset. Worth noting is that decommissioning has not been considered here—neither the NPV gains for deferred abandonment nor cost savings achievable by phased decommissioning (scaling down operating costs as parts of the infrastructure become redundant).
However, one of the challenges this kind of solution presents is the need for shared data and the trust that your joint-venture partner is not gaming its forecasts and cost structure to its benefit and your detriment. The data sharing side of things should be less of a technical challenge as we move increasingly toward cloud systems and a collaborative IT culture. However, the trust may be harder to shift. Sir Ian Wood is right to point out that greater collaboration in the North Sea is needed to maximise the value of what reserves it has left. The alternative is to leave money (and reserves) on the table. Perhaps it is worth considering how a more trusting business climate can be fostered as it is ironically profitable to establish.
1. The reason that Field 2 hypothetical COP is one year further out than that of the hub considered as a whole is to do with Field 1 opex being taken out of the equation in the hypothetical scenario where Field 2 can assume control of the Field 1 infrastructure without additional costs to Field 2.
3. That is, $65 per barrel for the entire operating period (2014 – COP)
4. To explain this further, consider that this is a Cost-Share Tariff mechanism. When the tariff levied is too low—less than or equal to $40/bbl in this case—Field 1 (and therefore the hub) will not stay open for any years in addition to the cost-share COP year, because it will not be profitable to do so for Field 1. Therefore the hub NPV is the same as it would be under the cost-share arrangement. Likewise if the tariff charged is too high (more than or equal to $70/bbl here) Field 2 will not stay open for any years after the cost-share COP date since it will not be profitable to do so—therefore The hub NPV will be the same for all tariffs higher than a certain value.
*This article was originally published for Palantir Economic Solutions (Dec 2014)