The Business Analytics team has recently undertaken a number of Asset Investment Planning (AIP) engagements with clients across the Utilities sector. Utilities companies are faced with the challenges of managing an ageing asset base (such as water mains, gas pipes, rail track) under increasing financial pressures. The utilities sector also faces significant regulatory scrutiny – for example, UK Water companies are currently approaching PR14, their price review for the next 5-year asset management period. We have been able to help clients to optimise their investment decisions so they can make the best use of their money to achieve tough performance targets.

But what is optimisation, and how does it vary from traditional, less analytically advanced planning methods? One of our software partners, Copperleaf, has put together the following example to illustrate the strength of investment optimisation:

Our fictitious client is developing a three-year asset management plan and has a portfolio of four potential investments. The cost profiles are shown in Figure 1, along with the value delivered by the investment which depends on which year the project commences (dependent on the nature of the company, the value may be financial, or it may be a composite risk reduction figure or performance benefits such as asset failures avoided)
Table of investment options
Figure 1: Investments under consideration

If the client has a budget constraint of just £3000 per year for the three years, can you find the optimal choice of investments?

I won’t launch into the answer immediately so you have to scroll down to see the results! In the meantime, I will consider what the choice would be using a more traditional approach to investment planning: prioritisation.

Prioritization of the options based on Total Value would be:     

  1. Investment 1     13200
  2. Investment 3     9000
  3. Investment 2     6000
  4. Investment 4     5000

Having prioritised Investment 1 first, it is then impossible to select Investment 3 (it would cost £4000 in the first year), so the achievable portfolio would be Investment 1 (starting year 1) and Investment 2 (starting year 2) TOTAL COST: £9000 TOTAL VALUE: 15200

Prioritization based on Value/£  (the best bang for your buck as my client said in Canada):          

  1. Investment 3     9000/£3000 = 3
  2. Investment 4     5000/£2000 = 2.5
  3. Investment 1     13200/£6000 = 2.2
  4. Investment 2     6000/£3000 = 2

In this case, you might prioritise Investment 3 &4 for starting year 1, then there is no budget to include Investment 1 but you can start investment 2 in year 2. TOTAL COST: £8000 TOTAL VALUE: 16000

You can see that this method gives differing results based on the prioritisation assumptions. Optimisation iterates through possible alternatives and picks the best combination. The possible combinations for this simple example are shown in Figure 2.
Table of potential plans
Figure 2: Optimisation candidates for evaluation

You can see that Candidate 1 is the solution we reached when prioritising by value and Candidate 4 is the solution we reached when prioritising by value/£. But the optimal solution is Candidate 3, which achieves the highest value despite not using the full £9000 budget in year 1. TOTAL COST: £8000 TOTAL VALUE: 18000.

This example considered just four investments but Utilities have to choose between thousands of different investment options. Stan Coleman, CTO of Copperleaf explains:

“Even with a relatively small number of investment alternatives and start dates, the number of possible combinations becomes overwhelming for a human to contemplate.  One of our customers was recently optimizing their plan to deal with a budget decrease and the optimizer presented an option to them that they initially found counter-intuitive – as it suggested delaying a large project that would bring a lot of value to the company. But the optimizer identified, and upon further reflection they agreed, that they could get more value by delaying the large project and continuing to execute a number of smaller projects. “

Real-life client issues are more complex and may involve additional considerations:

  • Thousands of potential investments
  • Multiple alternatives for each investment
  • Multiple cost constraints that vary over time (e.g. Capital and O&M)
  • Inflation, discounting, interest, loadings that all vary over time
  • Investment shifting by month
  • Longer timeframes (e.g. 30 years)
  • Resource constraints
  • Investment timing dependencies
  • Investment alternative dependencies
  • Mandatory investments (e.g. for legal reasons)

Kathleen McCorriston, Investment Planning Manager at Hydro One, one of our recent clients describes the benefits of optimisation:

“the investment planning process allows us to consistently compare investments across the organization by identifying the risk of achieving our corporate goals.  Optimization has helped us identify the best blend of investments to maximize the benefit and minimize risk to Hydro One and our Stakeholders”

The complexity of multiple factors in the planning process is one of the reasons that many of our clients are only now moving to optimisation instead of traditional methods.  Digital technologies and increasing computer power means it is now feasible to reach optimised investment portfolios swiftly and rapidly test the sensitivity of the investment plans to unpredictable future events such as extreme weather conditions and price fluctuations.  And optimised investment portfolios mean better use of the investment budgets leading to better service and lower bills for the consumer.