(733e) Scheduling of Distributed Chemical and Renewable Power Production with Regulated Energy Exchange

Authors: 
Allman, A., University of Minnesota, Twin Cities
Zachar, M., University of Minnesota
Daoutidis, P., University of Minnesota, Twin Cities
The coupling of distributed renewable power and chemical production can help to provide a use
for stranded renewable energy resources in rural areas, particularly if the chemicals produced will
be used in the same rural areas. However, such systems will still require interaction with the
power grid due to the stochastic nature of renewable production. It is well documented that
introducing additional renewable energy into the power market results in a required increase in the
ramp rates of fossil fuel generators and a risk of over or underproduction due to the stochasticity.
One way to mitigate these challenges is to introduce a regulatory market structure which obligates
renewable energy producers to explicitly limit the uncertainty and variability in their load by
providing a day-ahead commitment. Adjustable parameters can be used to specify how close
consecutive commitments must be to each other (mitigating variability), and how close the realized
energy exchange must be to the commitment value (mitigating uncertainty). Our previous work
has applied this regulatory structure to a system where an office building is powered by renewable
energy and shown that meeting commitments solely by adjusting power-producing or storage units
(i.e. microturbines or batteries) was quite expensive for renewable energy producers [1].

In this work, we apply our regulatory energy exchange framework to a system where ammonia
production is powered by wind energy, using electrolysis to generate hydrogen from water. Because
the power load for this system is a chemical process, there exists some inherent flexibility in the
load itself which can be optimized by scheduling. For example, ammonia production can be scaled
up or down, and electrolyzers can be turned on or off. The scheduling problem itself is formulated
as a mixed integer linear program that minimizes operating cost each hour using a 48-hour receding
horizon. Unit operating levels can change every hour, with the exception of the ammonia reactor,
which can change at most every 4 hours and must obey constraints in ramp rates. Power exchange
commitments are made 24 hours in advance, and we show that the flexibility in hydrogen, nitrogen,
and ammonia production can be used to avoid violating these commitments and paying penalties.
In addition, we show that the cost of regulating the energy exchange is minimal and that the
variability and uncertainty of energy exchange is significantly reduced.

We extend our work by analyzing how the power exchange variability and operating cost change
as various key design parameters are varied. We note that a tradeoff results from changing some
design parameters; for example, an ammonia plant which is too large will require more purchasing
of power, while one that is too small will make it difficult to meet power exchange commitments.
Other design parameters will have diminishing returns, such as the number of electrolyzers installed
or the amount of gas storage available. We also analyze the effect of location on operating cost and
establish that locations with a higher wind capacity factor can enable larger ammonia plants relative
to the installed wind capacity. This analysis enables better long term planning on where and how
to build renewable ammonia plants, and can be applied into a larger supply chain optimization
problem.

[1] Zachar, M. and Daoutidis, P. Microgrid/macrogrid energy exchange: A novel market struc-
ture and stochastic scheduling. IEEE Trans. Smart Grid, 8(2017), pp. 178-189.