Thermodynamics of Energy Recovery in SWRO Desalination Plants

Introduction and Background

Lack of access to potable water — or access in places where it is difficult to obtain — is a reality that many individuals encounter worldwide. Technology is capable of helping to improve this situation. An example of such technology is reverse osmosis, a procedure that utilizes membranes to separate salt from seawater. It applies pressure vessels that house three membranes, which are frequently spirally wound. Approximately 35 to 50% of potable water can be retrieved from seawater introduced into the desalination plant. Other essential components of the plant typically include an energy recovery system, a pre-treatment system, a post-treatment system, and pumps.

The operation of these desalination plants involves significant costs due to the energy consumed. Several approaches can minimize these costs: (1) combining the plant with other systems, (2) developing better membranes, (3) utilizing more efficient pumps, and (4) implementing new or enhanced energy recovery technologies. Commercially used energy recovery devices (ERDs) include the Pelton turbine, the pressure exchanger, and the turbocharger. It is worth noting that the Pelton turbine is perhaps the most widely utilized energy recovery device (Qureshi & Zubair, 2016).

Problem Statement and Significance

Implementation of pressure retarded osmosis (PRO) units as ERDs has yielded efficiencies approximately equal to or lower than that of the hydroturbine. For the range studied, PRO is not a feasible technique of energy recovery for reverse osmosis plants with seawater inlets, because limitations such as finite membrane area and concentration polarization would reduce performance even further.

Within the last five years, ERDs produced by the Swiss company Calder AG have been supplied to a number of the largest seawater reverse osmosis desalination plants around the globe, including facilities in Carboneras, Spain; Fujairah, United Arab Emirates; Las Palmas, Spain; and Tampa, USA. Approximately 90% of the brine reject energy is recovered by the ERDs in these facilities, leading to considerable savings in energy expenses. Calder's largest turbine model is rated at 1,800 kW, while its smallest model is rated at 20 kW. In large plants, ERDs are normally designed for a particular range or function. For instance, the turbine found in Trinidad — the largest train of its kind in the world — has been designed for 880 m³ per hour at pressures ranging between 39 barg and 72.4 barg, with a speed range varying between 2,235 rpm and 2,700 rpm. Close collaboration among the contractor, pump producer, and Calder helped achieve the best feasible efficiencies over the full range of performance, reaching no more than 2.5 kWh/m³ of product water.

Twenty years ago, energy recovery at SWRO plants was not broadly utilized, and where turbines were installed, they were based on reverse-operating pumps. The preferred choice of ERDs at that time was Francis turbines. Calder AG first evaluated the concept of an energy recovery turbine based on the Pelton wheel technology twenty years ago. The earliest prototype devices were based on standard hydro-electric impulse turbine hydraulics, but with considerable differences in material selection and construction requirements.

Calder energy recovery devices account for approximately 90% of the ERDs fitted to larger SWRO plants (in excess of 4,000 m³/day). Their hydrodynamic features permit a broad range of different operating parameters without considerable reduction in design efficiency. This means that pressure and flow variations in the RO desalination process do not significantly affect operating conditions or turbine efficiency. The impulse turbine functions across the entire range of working conditions. Cavitation does not occur within the defined working range, which guarantees an extended lifetime for the rotor system — normally adding up to ten years to its operational life. From a mechanical standpoint, the rotor is the only moving part in the system, meaning fewer components reduce the equipment's capital cost.

The reliability of the Calder energy recovery turbine is another significant factor. More than 1,000 units are currently operating globally, and reliability in excess of 99% has been demonstrated. In many regions where these products are in use, local communities depend entirely on fresh water from the desalination plant. A drop in production would immediately create serious problems. Furthermore, larger projects are frequently structured on a build-own-transfer or build-own-operate basis, so production directly impacts revenue and downtime may result in penalty claims. The most highly stressed components of the turbine are constructed from Super Duplex stainless steel, making them completely free from galvanic corrosion. Hydrodynamic element components — such as needles, wheels, and inlet nozzles — are cast and then machined to achieve high resistance to jet impact, high-velocity fluid friction, and cavitation.

The aim of this work is to assess the practical use of various PRO-based energy recovery devices for seawater inlets compared to established ERD technologies, utilizing exergetic efficiency alongside precise seawater characteristics. The research attempts to establish the performance of different energy recovery devices by quantifying the amount of energy required to operate the SWRO plant — that is, the power consumed in generating a given volume of water. This carries economic significance as it assists in determining the plant's overall economic and fiscal proficiency (Schneider).

Methodology and Assumptions

A SWRO desalination plant is assessed utilizing a number of ERDs. The utility of various PRO-based energy recovery devices for seawater feed is evaluated by comparing them with conventional ERD technologies using exergetic efficiency, which makes both economic and thermodynamic sense when accompanied by precise seawater characteristics.

An ordinary energy recovery device is compared to an energy recovery device utilizing pressure-induced osmosis. The configurations studied include:

Pressure exchanger (PX); pressure retarded osmosis unit coupled with hydroturbines (PRO-T); hydroturbine (T); throttling valve (TV); turbocharger (TC); pressure retarded osmosis unit combined with a hydroturbine and pressure exchanger (PRO-PX); and two-stage pressure retarded osmosis (2S-PRO-T).

The following assumptions govern the analysis:

There is insignificant pressure reduction in the ERD lines. There is no leakage at the pressure exchanger. PRO units have unit flow and counter-flow configuration. The impact of reverse salt diffusion and concentration polarization is ignored. The entire system is at a constant temperature. The feed water condition is regarded as the dead state: T₀ = 21.4 °C, P₀ = 101.325 kPa, and S₀ = 36.888 g/kg (except in cases where any of these parameters are varied). The efficiency of the turbocharger is taken as 70%, while the pressure exchanger is assumed to operate at 96% efficiency. The permeate salinity is taken as 0.4 g/kg. The SWRO plant is assumed to have a recovery ratio of 42%.

Conclusion

Various SWRO desalination plant configurations incorporating different ERDs are investigated using first- and second-law thermodynamic analysis. In addition to conventional ERDs, the analysis included single- and two-stage PRO configurations as well as PRO combined with a pressure exchanger. When the pressure exchanger was utilized as an ERD, the reverse osmosis plant performed best. The impacts of salinity, temperature, mass ratio, and pump and turbine isentropic efficiencies were all explored. Salinity was found to have the most significant impact on second-law efficiency. Additionally, where the PX was utilized, specific energy consumption was most strongly affected by pump isentropic efficiency, with an absolute variation of 0.4 kWh/m³.

For all elements studied within their respective ranges, every PRO-based configuration considered had exergetic efficiencies less than or approximately equal to the hydroturbine. Thus, PRO does not appear to be a suitable technique for energy recovery within the range studied. Future work should examine the impact of concentration polarization and finite membrane area more rigorously, as these real-world limitations are expected to further reduce the relative performance of PRO-based configurations.

References

Qureshi, B. A., & Zubair, S. M. (2016). Energy-exergy analysis of seawater reverse osmosis plants. Desalination, 385, 138–147.

Schneider, B. (n.d.). Turbines recover 90% energy in seawater reverse osmosis plants. Water World, 7(5). Retrieved from Water World.