For many months, I’ve struggled to model the economics of microgrids. For a set of paying customers of a homeowners’ association with 120 homes, for instance – a “load” – with daily and seasonal variations in demand, we can anticipate revenues easily enough.

Understanding microgrid costs, however, involves at least two complex optimizations followed by pro forma cash flow analysis. As best I can tell, this has not been done.

A microgrid typically has multiple generation sources and a battery. Some people call such systems “hybrid,” but that states the obvious, physical truth. They ought instead be described as integrated systems, since their working combination should be more than additive, and follow a well-articulated logic toward solving an objective function, for instance, production at the lowest cost; with the lowest emissions; continuous operation with quality electricity; or predictable demand with low peaks.

Consider a 100,000 DWT container ship, a sizeable floating microgrid. Its propulsion requires ~ 70 MW engine power generated by burning heavy fuel oil. In addition, it has sizeable power systems for cargo refrigeration; air-conditioning for the crew’s living spaces; critical communications systems; and other loads much like any community, likely about 15 MW. Ships do not emphasize renewable generation, yet like any terrestrial microgrid, they have batteries, diesel generators, and waste heat recovery systems, optimized for cost savings and reliability. Do ships offer lessons for microgrid design? Shipbuilders might design optimized microgrids.

Two Optimizations and a Discounted Cash Flow

Capacity Planning: The first challenge of microgrid design is: For the given demand, in what combination should we deploy, say, solar, wind, diesel or gas generator, and batteries? How many kW of each? This sizing is a thorny challenge. Every location, having different resource endowments, needs customization. Homer Software claims to address this important issue, thereby permitting us to budget for capital costs. How many solar panels are needed for the sunlight available in the proposed location? What are the wind turbine sizes, and their number, for the given topography? How much battery storage, and the size of generators needed, to meet the given load?

When the capital needed for the various generation sources is represented as payments spread over a number of years, it sets the first constraint on aggregate economics. The revenue estimated has to cover that, plus operating costs.

Operations Modeling: The second constraint is efficient operations. How should we utilize the deployed resources in an optimum manner to reduce overall operating costs? How should we maximize solar use, reduce diesel or gas use, store electricity generated in batteries, and use the stored energy after sundown? This is an integer-programming problem; a number of recent research papers have made admirable contributions toward addressing it.

To manage this optimization well, however, we need a good control network interfacing with the resources, a non-trivial engineering and design challenge, for the resources must satisfy a dynamic demand. This likely requires experimental prototypes working simultaneously with associated analytics.

Financial Modeling: Beyond capacity planning or sizing, and optimized operations, we need to make assumptions about the differing life of solar panels and batteries; the cost of money – the weighted average cost of capital (WACC) – for the business undertaking the project; factor in risk through a discount rate; and assume a terminal multiplier at the end of, say, ten years in the cash flow statement. We need to anticipate operating costs as in any business plan – for people, offices, sales, maintenance, fuel consumption, and the like.

The goal is to assess the economic feasibility of microgrids at a high level, and understand the relative role of the various economic drivers. For simplicity, we assume mature technologies, a standalone microgrid, and no subsidies in our estimates. These factors may be included in future iterations of the model.

At this stage, we are ready to calculate the Net Present Value of the project, and conduct breakeven analysis for the system. Is the project worth doing? Now? If not now, when? What are the sensitivities?

Environmental Barriers To Realizing Microgrids

Mind you, these issues are merely the technical challenges of establishing a business case. Regulatory and public policy issues need to be addressed too. For instance, in the U.S., does the microgrid service territory cross public rights of way? If it does, do microgrid operators violate the franchise rights of incumbent utilities?

If the microgrid expects to connect with the macrogrid, there are the additional equipment costs, and operating challenges of islanding. Let us assume that they are addressed satisfactorily in the IEEE 1547 standard, and through global manufacturers.

We next need to bridge historic academic boundaries, and their reflection in any organization or team. Electrical engineering, operations research, and public policy, among other disciplines, are distinct competencies with well-defined boundaries that are not easily breached. Cross-disciplinary research can be career limiting in that it belongs neither here nor there, and requires the burden of new learning.