At last count, Somerset’s Hinkley Point C nuclear power plant will cost ~£22.5 billion(bn) to build. The original estimate of ~£6.5 bn seems a long time ago. Such dramatic cost overruns are a sadly familiar feature of major infrastructure projects but are just one of the issues facing those seeking accurate valuations for insurance reporting in the power sector.
With construction costs typically falling between 500 – 8,000 $/kW (depending on type, size and location), and a standard construction time ranging from one to six years, building a power plant is a complex process. There will be multiple suppliers of equipment, materials and labour, sourced both domestically and internationally. The main factors driving construction costs are: (1) cost of key commodities: steel, copper, concrete, etc., (2) a limited number of critical equipment manufacturers and suppliers, (3) a shortage of skilled labour, and (4) tariffs on imported materials and technology in many jurisdictions.
Key Considerations
Who is the Likely Supplier?
While major pieces of equipment such as turbines and generators have a limited number of global suppliers, other assets such as boilers might be supplied locally or from a specific international manufacturer, depending on the project specifications. General equipment, such as tanks, vessels, pumps and compressors, are typically available from local suppliers.
What is the likely installation scenario?
What Is the Likely Installation Scenario?
Will the power plant be erected using predominantly local labour and equipment suppliers - which might be the case in developed countries? Or will there be a combination of local and ex-patriate labour (A typical situation in emerging economies with a less developed industrial base)? It is crucial to understand the potential differences in labour costs.
Additionally, international equipment suppliers may rely on local labour supervised by the supplier’s own engineering staff, or simply employ their own installation team to deliver a turnkey solution. However, more general equipment will likely be installed by local teams.
A similar situation exists with real estate assets, where local labour is usually used to construct standard buildings, while specialist international labour might be necessary for buildings that are required to meet specific parameters, or when using non-standard materials and construction processes.
What Local Policies and Regulations Are Applicable?
Power generation is one of the central issues in the national energy policy but is also influenced by global trends. For example, the burgeoning commitment to carbon neutrality, with accompanying divestments from fossil fuels, will likely affect not only the availability but also the import duties on a specific technology.
Which Valuation Method is Most Appropriate?
There are three methods that can be used to estimate the reinstatement cost of a facility; each with advantages and disadvantages.
Direct Costing Method
Direct costing is generally considered the most accurate method for replacement cost development. The costs of new assets are typically acquired directly from the original manufacturers, or from recent price quotations, industry publications or price catalogues. These costs are then adjusted for any direct and indirect installation costs.
The major challenge with the direct costing method is the availability of data. For some assets, current prices may not be available (for example, an item of equipment that is no longer manufactured and/or the manufacturer no longer exists). For others, the original manufacturer may not be willing to disclose the current prices to a third party.
Moreover, direct and indirect installation costs are rarely available, as they are specific to the location and type of power plant.
Trending Method
The trending technique is perhaps the most commonly used approach owing to its relative simplicity. It is based on the original cost new of the asset to the first owner. This historical acquisition cost is typically recorded in the company asset register and should include freight, tax and installation—all costs incurred to make the facility operational whether funded by the insured or through government incentives. It is then adjusted (multiplied) by a specific price index respective to the time of the original purchase to derive a current replacement cost for the asset.
However, as trend factors are usually based on undisclosed data, it is difficult to gauge the level of accuracy. Also, while there are composite and general indices that may be applied to broad spectrums of assets, it is better to use an index based on the prices of directly comparable assets.
The trending method is generally applicable and provides a reliable indication in the following cases:
- The subject property is relatively new: The rate of inflation and technological advances will affect the time period over which assets can be reliably trended.
- The subject property is located in a stable economy: Current or past hyper-inflation in the subject country precludes the use of the trending method as such volatile development of prices makes any data statistically unreliable.
- Accurate records are available: For the trending method to be robust, the company needs to maintain accurate records. In particular, the true historic cost and acquisition date of assets must be available. Where assets have been purchased secondhand or transferred from affiliated companies at their net book value, the trending method is considered unreliable.
- Historic cost overruns have been passed on to the client: If cost overruns are paid for by the contractor and not passed on to the client, the trending method may omit an often very significant portion of the overall project cost.
The main advantage of the trending method is that it includes all installation costs specific to the subject plant, which are normally recorded in the fixed asset register. Adjustments may be necessary, however, in a number of situations. These include prior discounts, capitalized maintenance costs and/or capitalized indirect costs not covered by insurance policy. These include site preparation, interest and finance charges, design and engineering, licenses, permits or operator training. If an asset was purchased from an international manufacturer, then it may also be necessary to make adjustments for currency fluctuations between the original acquisition date and the appraisal date.
Under this approach, the concluded replacement cost new is based on an exact replica, with no allowance for technological progress. As such, the trending method may not be fully applicable for older assets or where there has been rapid technological development.
Benchmarking Method
The benchmarking method is based on known prices of assets with similar physical characteristics, functionality and utility.
For plant and equipment where the price is not known, but figures are available for units with the same functionality but different capacity, a cost-to-capacity formula can be used (Jelen, F. C., Black, J. H., “Cost and Optimization Engineering,” McGraw Hill, 1983).
Benchmarking can be applied at an individual level, in some cases at a process level, or even at a battery limits level. A battery limits approach benchmarks the subject property against the total investment needed to construct a production plant producing specific products at given capacity. This approach is practical for power plants, refineries, chemical and other process plants.
However, cost new estimation derived from the use of “rules of thumb” should not be given substantial weight. This approach may be used as a quick cross-check of the cost new derived by other methods, or when doing a sensitivity analysis where a high degree of accuracy is not required.
What other factors need to be considered?
One of the most significant issues impacting the power sector is cost overruns.
Looking at historic examples from the 1970s, the costs of building new nuclear power plants began to spiral out of control—actual costs were two to three times higher than estimated during licensing or at the start of construction.
Indeed, the nuclear industry has a very poor track record in predicting plant construction costs and avoiding cost overruns. As shown in a study by the Department of Energy, the actual costs of 75 nuclear power plants in the U.S. exceeded initial estimates by more than 200% (Schlissel, D., Biewald, B., “Nuclear power plant construction costs,” Synapse Energy Economics, 2008).
Another more recent analysis of cost overruns looked at 400 energy infrastructure projects, concluding that the average cost overrun per project was almost $1 bn, which represented an average overrun of 66%:
| Power plant type | Average project overrun |
| Hydroelectric |
70.6% |
| Nuclear |
117.3% |
| Thermal | 12.6% |
| Wind farm | 35.0% |
| Solar facility | 39.0% |
Source: Sovacool, B. K., Gilbert, A., Nugent, D., ‘An international comparative assessment of construction cost overruns for electricity infrastructure’, Elsevier, 2014
In addition, recent analysis of Brazilian hydroelectric dams calculates an average cost overrun of 97.5% (Callegari, C., Szklo, A., Schaeffer, R., “Cost overruns and delays in energy megaprojects: How big is big enough?”, Elsevier, 2018). As seen with the UK’s Hinkley Point C, cost overruns are certainly still with us. So, how can the cost overrun be accounted for in a valuation?
Using the trending technique (if the cost overruns have been fully passed on), the replacement cost new should capture the project overrun costs in the same proportion as incurred during original construction.
When using other methods, an assessment of potential cost overruns is more difficult and may rely on the application of a contingency based on market observations. In all cases, the valuation report should clearly state whether the potential for cost overruns is included, and how it was determined, so that the insurer can make adjustments if needed.
Conclusion
The determination of supportable replacement costs in the power sector is never straightforward. It requires deep sector experience and a thorough understanding of the available data if serious inaccuracies in the replacement cost estimate are to be avoided.
