Hybrid Energy Systems

A traditional design consists of making assumptions about the future, and then designing a system to fit with those assumptions. There are a couple of areas for improvement in this methodology. For one, the future is inherently uncertain, and it is typical for designers to use an average value for various uncertainties. The design is then based off of this average value. The problem with this is that designs based on average values are likely to be wrong. Not taking into account that uncertainties will often be associated with a probabilistic distribution can cause decision makers to have unrealistic expectations about how risky a project will be.

Sometimes, designing based on an average can actually cause reality to be strictly worse than the average. Consider a natural gas plant. There is an expectation that the demand over the next 20 years will be 140 MMcf for this plant, and thus the capacity is designed around this number. The problem is, for periods where the demand is above 140 MMcf, the plant is unable to produce more, but the plant still produces less when the demand is less than 140 MMcf. Designing around the average almost guarantees that it operates at less than the average quantity. Considering the uncertainty as a distribution allows for a more realistic look into the viability of a design.

Considering the uncertainty as a probabilistic distribution can allow a decision maker to make more accurate estimations of the value of a project, but can the value of the design process be increased? One potential alternative to fully designing a system at the outset, is to actually allow for the system to be evolvable. This evolvability can allow the system to adapt to changes in the environment, and potentially allow for more value to be captured. The evolvability can be though of as future decisions that a decision maker would make on a project with a long life. A power plant that is in operation for 60+ years does not remain static, and in fact is improved or expanded upon as needed. Not taking these decisions into account ignores an aspect of the design that can be significant. The difficulty with attempting to account for future decisions, is that each decision should be the optimal decision. Therefore, a problem with future decisions invovlves optimizing the future decisions within an optimization for the base design. There is also the issue of when to perform these future decisions. For larger decisions, it may not be necessary to evaluate the design everyday, but every year. Even so, for a project that lasts 60 years, that involves 60 optimizations per base design, which is also being optimized. These nested optimizations can bring the computational complexity to such a point that it is not pragmatic. Methods are therefore being investigated to make the design process more valuable by utilizing this evolvability, called Real Options Theory (ROT), with Value-Driven Design (VDD), and by investigating how to increase the efficiency of predicting value through various methods.

Case Study: Hybrid Energy Systems

In order to test this hypothesis, a case study involving a Hybrid Energy System (HES) is in progress. An example of an HES is shown below, for the case of a multiple-input, multiple-output power plant. In the case study, a nuclear plant is capable of producing electricity, or diverting steam to the chemical plant for the production of transport fuels. The plant diverts steam based on the quantity of power being generated by the renewable wind farm. This allows the base load nuclear to always be producing steam, as cycling a nuclear plant is expensive, and still allow for a level of load following. In this case, the chemical plant helps to act as a storage unit for excess energy produced by the wind farm. The HES is a multidisciplinary complex system. The field of power generation is also a very uncertain one, which relies upon atmospheric conditions (temperature, wind speed, and humidity), future technologies (batteries), as well as various prices of both fuels (natural gas) and products (electricity and transport fuels).