A McGill University-led research team has demonstrated the feasibility of a sustainable and cost-effective way to desalinate seawater. The method—thermally driven reverse osmosis (TDRO)—uses a piston-based system powered by low-grade heat from solar thermal, geothermal heat and other sources of renewable energy to produce fresh water.

Though previous research showed promise, this study is the first to analyze TDRO’s thermodynamic limits. The results have brought researchers closer to realizing the technology which could improve access to water and increase the sustainability of infrastructure.

“Most desalination is done by reverse osmosis, which uses electricity to drive water through a membrane,” said Jonathan Maisonneuve, study co-author and Associate Professor of Bioresource Engineering.

“The challenge with using heat is that it takes a lot of it to do what you could with a little bit of electricity. So, if we can find a way to harness existing heat from renewable sources, that could be very advantageous, because it’s so abundant,” Maisonneuve said.

“Thermally driven reverse osmosis: thermodynamics of a novel process that uses heat for desalination and water purification,” by Saber Khanmohammadi, Sanjana Yagnambhatt, Dan DelVescovo and Jonathan Maisonneuve, was published in Desalination on Oct. 15, 2025.

Addressing the water and energy crises

Electricity-based desalination, which is often inaccessible in remote areas, requires about one to four kilowatt hours (kWh) to produce one cubic meter of fresh water.

According to the researchers’ analysis, which optimized several elements of a design proposed by MIT researcher Peter Godart, TDRO would require 20 kWh per cubic meter.

“There’s still a big difference when you compare it to one to four kWh, but because heat is cheaper than electricity, we don’t have to totally close that gap,” Maisonneuve said.

TDRO works by heating and cooling a small amount of fluid in a sealed chamber, known as the working fluid. This temperature fluctuation expands the working fluid, causing it to drive a piston to push seawater through a reverse osmosis membrane—effectively combining a thermodynamic cycle with water purification.

By studying and optimizing the ratio of working fluid to seawater fluid, as well as the piston sizes, the researchers demonstrated that TDRO has better performance potential than previously thought. The method also compares well against existing thermal desalination technologies, but they say further study is required.

“Next, we need to model it in detail, see how quickly the system can operate and introduce a number of non-ideal effects, such as heat loss through the environment,” Maisonneuve said.

Using a network of synchronized sensors, a new system provides energy and meteorological data every tenth of a second to more accurately predict the performance of solar plants.

Solar plants do not always provide energy. Rather, they do so only when they can, as they depend on the weather. A single cloud, for example, can cause a drop in production.

To address the uncertainty inherent in solar plant operations, the Instrumentation and Industrial Electronics research group at the University of Cordoba has developed a new monitoring system that uses a variety of sensors to provide more detailed, precise, and synchronized information about solar plant performance. This is a Big Brother-style monitoring system designed not only to observe and predict these plants’ performance, but also to adjust the auction that sets energy prices to more realistic conditions.

The research is published in the IEEE Sensors Journal.

The goal was to analyze how solar plants generate energy in relation to weather conditions. To achieve this, the team utilized devices that have been on the market for several years, to which they applied new implementations. Thanks to these devices, the collection of equipment, which they have named Extended Phasor Measurement Units, can gather energy (such as current, voltage, and frequency) and meteorological (such as solar radiation) data.

The key is that this data is collected every tenth of a second, in great detail; so detailed, in fact, that they generate around 2 to 3 gigabytes of data each month. Furthermore, because the devices are synchronized with each other and located in different areas, including nearby solar plants, the information they provide allows us to see what has happened, and anticipate what is going to happen.

As UCO researcher Víctor Pallares López explained, “It’s about closely monitoring the systems to gather as much information as possible and being able to react to any negative effects that could affect the stability of the electrical grid.”

The more closely they are monitored, the sooner action can be taken to isolate the system from potential disturbances by disconnecting it and preventing their spread.

The sensors have been fine-tuned, tested in both the laboratory and at two Pozoblanco plants, and all the errors have been reviewed, so the team is now immersed in the second phase of the project: analyzing all the data generated.

This work is currently being conducted as part of the national project “Edge management of photovoltaic plants based on an analytical architecture with near-perfect temporal precision,” with reference PID2024-158091OB-C21.