It was a dark and stormy night as we sat around the conference table discussing the best way to arrange the grid connection of the company’s new wind farm that was in design. Everyone was friendly, but it was a feisty conversation. It wasn’t the first wind farm the company had developed, and since the first wind farm had been built, the company had formed a reliability engineering department. The old-timers wanted to replicate the existing design, whereas the reliability engineers wanted to consider availability, risk tolerance and failure rates to have a design that ensures the lowest total cost with the greatest reliability. Opinions varied, but boiled down to two options: a straight bus connection and a dedicated spare transformer for improved reliability, and a ring bus that allowed for added redundancy and kept power output in the event of the most common failure modes. The reliability engineers decided to settle the debate through a lifecycle analysis of varying designs.

Using data from the existing ten-year-old wind farm, capacity limitations, equipment failure rates, downtime, repair costs, spares lead times, storage costs, power operations, and power value projections were used to build financial models. Bids for installing the straight bus and ring bus designs were calculated and costs to purchase, install, store and maintain the spare transformer were collected. Some assumptions had to be made, to which everyone agreed. A technician spits out, “I don’t get it, we know what works, let’s do the same design as we have,” during a discussion of how to maintain the transformer in each of the designs to minimize downtime in the event of failure. “Well, that is the power of designing with the end in mind. The company knows what level of availability it wants and what risks it is willing to accept, so it is up to the design to meet those requirements,” said the reliability engineer.

So, rather than simply considering the capital costs of purchase and installation, the team used Monte Carlo simulation to look beyond the capital investment and consider the total cost of ownership for five years and eight months, the period before selling off the farm to another utility. Running one thousand simulations over the twenty-year period, the results were obvious to those funding the project with a payback of eight months. When the recommended design was presented, the technician stated, “Oh, I hadn’t thought about risk reduction and cost avoidance, this makes a lot of sense.”