(244d) Optimal Pump Network Reconfiguration Scheduling Considering Time-of-Use Electricity Prices | AIChE

(244d) Optimal Pump Network Reconfiguration Scheduling Considering Time-of-Use Electricity Prices


El-Farra, N. H. - Presenter, University of California, Davis
Palazoglu, A. - Presenter, University of California, Davis

The electricity demand over the day is subject to significant fluctuations, with considerably larger demand during peak hours than off-peak [1]. An increasing penetration of intermittent renewable energy generation further increases the variations, and therefore the costs for the utilities to meet the demand. One approach to deal with this challenge is Demand Response (DR) which aims at changing the electricity end-user’s demand using penalties or incentives such as variable electricity prices [2, 3].

While real time electricity pricing is only implemented in few applications so far [2], a simple way of realizing a DR strategy, Time-of-Use (TOU) rates, is widely adopted. For TOU electricity pricing, the utility sets several electricity price levels per day. The prices are fixed and are only adapted in larger time scales. A higher electricity price during peak hour clips the demand and shifts part of the demand to off-peak hours [3]. TOU plans are common for commercial and industrial consumers, but they are also implemented for residential consumers and offer an opportunity for the consumer to save costs by changing the consumption pattern according to the electricity prices.

Pump networks are widely used in many industries. According to the European Commission, pumps are the largest consumers of electrical energy in the EU [4, 5]. Especially in the paper and pulp industry or in water distribution [6], where large-scale pump networks are applied, the electricity costs incurred by pump networks are significant. Therefore, cost saving potential in pump networks is large [4, 5].

An optimal configuration of pumps can lower electricity consumption significantly [5, 4]. However, the operation of the most energy efficient configuration may not always be desirable from an economic point of view, since a higher number of pumps or technologically more elaborate pumps may also entail higher maintenance costs. Therefore, the optimal configuration, which has to consider both electricity and maintenance costs, may depend on the time of day and the electricity price levels of the TOU plan and has to be found in order to minimize operational costs.

In this work, we propose to reformulate the traditional static pump network optimization problem with consideration of TOU electricity prices and maintenance cost based on the case study in [6], to find the optimal pump network scheduling during a day. The focus of the reformulated problem is on the optimal operation of an existing network superstructure, in order to eliminate the direct consideration of the capital costs. However, the maintenance cost is correlated to the capital cost. The proposed methodology is applied to a pump network superstructure consisting of three layers, each of which is occupied by a different pump type with unique pump equations. Variable speed centrifugal pumps are assumed to allow for a change of rotational speed in each configuration. The required volume flow and pressure gradient over the pump network are assumed constant. The optimum pump network configurations for a fixed interval of electricity prices are determined for varying plant sizes by solving the reformulated MINLP-problem for global optimal solutions using the Branch-and-Reduce Optimization Navigator (BARON). It is shown that switching between different configurations can lead to overall cost savings. It is also found that the electricity prices for which a transition between two configurations is economically justifiable varies greatly with the plant size. A case with specific parameters could be found for which a reconfiguration is optimal during the peak-hours of a real TOU electricity price structure. 


[1] M. H. Albadi and E. F. El-Saadany, "Demand Response in Electric Markets: An Overview," in Power Engineering Society General Meeting, 2007. IEEE, Tampa, FL, 2007.

[2] I. Harjunkoski, X. Feng and E. Scholtz, "Smart Grid: Industry Meets the Smart Grid," CEP, pp. 45-50, 08 2014.

[3] C. W. Gellings, "The Concept of Demand-Side Management for Electric Utilities," in Proceedings of the IEEE, Vol 73, No 10., 1985.

[4] European Commission; ETSU, AEAT PLC, Study on Improving the Energy Efficiency of Pumps, 2001.

[5] Hydraulic Institute, "Pump Life Cycle Analysis - A Guide to LCC Analysis for Pumping Systems - Executive Summary," 2001.

[6] T. Westerlund, F. Pettersson and I. E. Grossmann, "Optimization of Pump Configurations as a MINLP Problem," Computers & Chemical Engineering, pp. 845-858, 09 1994.