Electrical Power from Sea and River Water by Reverse Electrodialysis: A First Step from the Laboratory to a Real Power Plant

Abstract

Electricity can be produced directly with reverse electrodialysis (RED) from the reversible mixing of two solutions of different salinity, for example, sea and river water. The literature published so far on RED was based on experiments with relatively small stacks with cell dimensions less than 10 × 10 cm(2). For the implementation of the RED technique, it is necessary to know the challenges associated with a larger system. In the present study we show the performance of a scaled-up RED stack, equipped with 50 cells, each measuring 25 × 75 cm(2). A single cell consists of an AEM (anion exchange membrane) and a CEM (cation exchange membrane) and therefore, the total active membrane area in the stack is 18.75 m(2). This is the largest dimension of a reverse electrodialysis stack published so far. By comparing the performance of this stack with a small stack (10 × 10 cm(2), 50 cells) it was found that the key performance parameter to maximal power density is the hydrodynamic design of the stack. The power densities of the different stacks depend on the residence time of the fluids in the stack. For the large stack this was negatively affected by the increased hydrodynamic losses due to the longer flow path. It was also found that the large stack generated more power when the sea and river water were flowing in co-current operation. Co-current flow has other advantages, the local pressure differences between sea and river water compartments are low, hence preventing leakage around the internal manifolds and through pinholes in the membranes. Low pressure differences also enable the use of very thin membranes (with low electrical resistance) as well as very open spacers (with low hydrodynamic losses) in the future. Moreover, we showed that the use of segmented electrodes increase the power output by 11%.

Energy Recovery from Solutions with Different Salinities Based on Swelling and Shrinking of Hydrogels

“…32 Assuming membranes used in PRO are 0.1 mm thick, with a specific weight of about 1 g/mL, 1 m 2 of membrane would be 100 g. Thus, energy recovery rates from PRO would range from 10 mW/g to 27 mW/g membranes. For RED stacks, assuming 1 W/m 2 that has been produced in both small-and large-scale systems, 10,11,33 this translates to 10 mW/g membranes. However, cost of the hydrogels must also be considered in this comparison with membrane-based technologies.…”

Pressure drop at low Reynolds numbers in woven-spacer-filled channels for membrane processes: CFD prediction and experimental validation

“…2 of [13]). Therefore, pressure drop due to the manifolds may be high [6,11,13,14,18,24,25]. Nevertheless, an improved geometry of the distribution system can dramatically reduce this contribution, while the increment of pressure drop along the channel due to a spacer remains quite higher with respect to the empty channel, as clearly shown in [14], and the spacer features have a decisive impact on pressure drop [7,14,20].…”

“…Therefore, in the optimization of the stack and of the operating conditions one has to account for the combined effects of several factors on various aspects [15][16][17]. However, among these, pressure drops and consequent power consumption might have the major impact on the process efficiency [7,11,13,14,18] especially when highly concentrated solutions are employed [19].…”

“…In general, it can be divided in: (i) the pressure drop in the distribution/collection system and (ii) the pressure drop along the channels [18].…”

“…The channel thickness is usually ~100-500 μm, the channel length ranges from 10 cm (laboratory scale) to ~40 cm (prototype scale); the typical linear flow velocities at which the maximum net power is obtained are around 1 cm s -1 , corresponding to very low Reynolds numbers (lower than 5) [7,[10][11][12][13][14]18,20,21]. Therefore, the fluid flow regime is perfectly steady.…”