Predictive power management strategies for
stand-alone hydrogen systems: Operational impact
Adel Brka
*
, Yasir M. Al-Abdeli, Ganesh Kothapalli
School of Engineering, Edith Cowan University, Joondalup WA 6027, Australia
article info
Article history:
Received 12 September 2015
Received in revised form
28 February 2016
Accepted 13 March 2016
Available online xxx
Keywords:
Hydrogen
Predictive
Power management strategy
Renewables
Device intermittency
Optimisation
abstract
This paper compares the operational impacts of both predictive and reactive Power Man-
agement Strategies (P-PMS and R-PMS). The study is implemented for a stand-alone hybrid
system based on wind turbines (WG), batteries (BAT) and hydrogen technology. The P-PMS
uses real-time Neural Network (NN) predictions of wind speed and load demand to adjust
the control set points affecting the switching of devices. The study also analyses the effects
of using another intelligent technique, Particle Swarm Optimisation (PSO), for real-time
optimisation of fuel cell operation. Genetic Algorithms (GA) are used to optimally size
the hydrogen system. The methods employed include MATLAB simulations to implement
the three intelligent techniques (GA, NN and PSO) and integration of experimentally
derived fuel cell characteristics as well as highly dynamic electric load and wind speed
profiles. The research presented in this study is an extension of an earlier work in which
the concept of P-PMS was experimentally validated and the effects of some software and
hardware related controlling parameters assessed. This paper however goes further by
analysing the impact of using P-PMS on the economic and operational characteristics of
stand-alone hydrogen systems by benchmarking it against an R-PMS. Results reveal that a
hybrid system operating under a P-PMS outperforms that with an R-PMS in terms of cost,
renewables penetration and environmental footprint. In addition, this study showed that
P-PMS can help mitigate the impact of the transient response of the backup components
(fuel cell and electrolyser) on the system sizing and operation. However, these merits are
realised only if a particularly high reliability of load satisfaction is required. Results also
show that a P-PMS highly depends on the accuracy of the employed (NN) prediction tool.
The proposed predictive strategies are proven to be better than other solutions that exist in
literature in terms of reducing the cost.
Copyright © 2016, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights
reserved.
Introduction
Solar-PV and wind energy systems are hybridised by adding
backup prime movers and energy storage media so as to
reliably meet electric power even during higher demand pe-
riods [1]. Diesel generators [2] and batteries [3] are traditionally
used in this context but these are expensive to operate and
maintain at remote locations [4], as well as having undesirable
environmental impact [5]. Batteries can also suffer from short
* Corresponding author. School of Engineering, Edith Cowan University, 270 Joondalup Drive, WA 6027, Australia. Tel.: þ61 432611745.
E-mail addresses: abrka@our.ecu.edu.au, adilbrka1@gmail.com (A. Brka).
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: www.elsevier.com/locate/he
international journal of hydrogen energy xxx (2016) 1 e14
http://dx.doi.org/10.1016/j.ijhydene.2016.03.085
0360-3199/Copyright © 2016, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Brka A, et al., Predictive power management strategies for stand-alone hydrogen systems: Opera-
tional impact, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.03.085