A deep dive into the evolution of pumped-hydro storage technology

Print Friendly, PDF & Email

How hydro technology defied gravity and became the gold standard for grid-scale energy storage.

Print Friendly, PDF & Email

The technology behind hydro-electric generation has come a long way since it simply went with the flow of water running downhill.

The development of reversible turbines in the 1920s was the first significant milestone in a century of innovation leading to the sophisticated grid-supporting capabilities of pumped-hydro storage systems now operating in Switzerland and proposed for other major projects worldwide.

Turning the tide charges the battery

“The first reversible turbines were axial” (water flowed parallel to the shaft), says Pierre Leroy, global hydraulic R&D director for GE Renewable Energy.

With the ability to not only generate electricity by harnessing downhill water flow, but to switch direction and use electricity to pump water from a lower reservoir to an upper reservoir, reversible turbines allowed elevated bodies of water to be used as ever-renewable energy-storage facilities.

Early pumped-hydro storage projects were restricted by turbine capacity to “low head” installations — “head” being the term used to describe the difference in height between the two reservoirs.

In the US, for example, the country’s first pumped-hydro storage project installed in 1930 by the Connecticut Electric Light and Power Company, had a head of 70 metres.

With hydroelectricity output proportional to the head times the water volume, the challenge became how to design turbines that could mechanically cope with higher heads and flows.

Heady flows of data minimise bad vibrations

In Switzerland, the Kraftwerk Linth-Limmern project has a head of 630 metres between the upper reservoir, Lake Mutt, and the lower Lake Limmern. The innovations that enabled such a power-boosting increase in elevation have been both physical and computer generated.

Introduction of the centrifugal pump-turbine in 1848 put a new spin on water’s energy-producing potential. Still known today as the Francis turbine after its American developer, James Francis, the technology improved on the efficiency of axial designs by accelerating water in a radial direction to extract more energy.

Francis turbines could also be designed to precisely match the water flow and pressure from a large range of heads. Now produced by several manufacturers, Francis turbines are the most widely deployed hydro turbines because of their versatility and efficiency.

“The combination of mechanical design and of hydraulic calculation allowed us to get more and more output,” says Leroy of his work in developing the Francis turbine for GE Renewable Energy.

Modelling known as computational fluid dynamics (CFD) is central to optimising turbine design for increased output while maintaining machine integrity and reliability. But even models using only basic reference points involve a deluge of data.

The scale of GE global operations and its dedication to driving productivity using Predix, the GE operating system designed for industrial applications, gives any of the company’s dedicated research centres access to a cumulonimbus of cloud-based computing capacity.

“Simulations showing machine behaviour under different conditions of flow and turbulence can run for several days,” says Leroy, “Now they can include very specific details that we were previously unable to capture.

Increases in computing capability have allowed us to more accurately simulate the effects of different heads and flows on the phenomena occurring inside our machines.”

Among the results of such precise computer modelling is a 3.5% increase in GE turbineefficiency over the past 45 years. “We are now achieving 92-94% efficiency, which means more power to the grid and more efficient use of the energy available in both pumping and generating modes,” says Leroy.

Digital modelling has also allowed GE Renewable Energy to extend the effective range of operation of its hydro turbines.

Where turbines of the past were designed for set points of operation, and any significant deviation from these points created mechanical and hydraulic problems such as vibration and cavitation (formation of masses of bubbles around the impeller that explode causing damaging shock waves), GE Renewable Energy turbines have since been engineered to operate flexibly over a range of conditions without exhibiting such damaging phenomena.

Revolutionising responsiveness to grid conditions

By far the greatest technological leap of recent years has been in the design of variable-speed pumped-hydro storage plants.

The intermittent and fluctuating output of energy sources such as solar and wind has driven research into hydro plants that can rapidly respond to store excess energy as efficiently as possible, or dispatch energy to meet changes in consumer demand, while simultaneously maintaining power frequency (50 Hertz in Australia) within the grid whatever sudden changes in input or uptake occur.

“For more than 100 years, the electrical machines used for energy production have been synchronous motor generators — they always ran at the same speed, synchronous with grid frequency,” says Alexander Schwery, Hydro chief electrical engineer for GE Renewable Energy.

“Now they need to respond to changes in power input from renewables while they are in pump mode, to maximise the uptake of excess power in the grid. But frequency regulation in pump mode is not possible with single-speed equipment.”

Over the past two decades hydro-equipment manufacturers have imagined a variety of technologies to overcome this limitation, including hydraulic short circuits and full converter-coupled synchronous machines, but they also envisaged and developed a new type of machine.

Careful consideration of efficiency, the cost of hydro-storage civil works and equipment, led to the idea of double-fed induction machines for large storage projects. In these machines the grid frequency and the mechanical speed of the machine are decoupled, enabling the machine to change speed and consequently change power in pump mode.

Compared to the straightforward process of designing hydro plants around synchronous technology in which, “The hydraulic team designs a system according to head height and flow, and defines speed and power output for the electrical team, ” Schwery says designing variable-speed systems requires great interdisciplinary interaction.

“To be asked to come up with systems that provide optimal flexibility as well as performance is an engineer’s playground — it’s fantastic to be there,” says Schwery. He adds, “The interfaces between turbine, generator and converter are complex, and the design process becomes a matter of iteration.

We negotiate requirements and specifications between the hydraulic, generator and converter teams, closing the loop many times to reach the ideal solution in terms of size, cost and capability for each installation.”

Late last year, GE Renewable Energy commissioned the most technologically sophisticated variable-speed pumped-hydro storage plantto date, in Switzerland’s Linth Valley.

The Linth-Limmern plant’s vast underground cavern between two alpine reservoirs houses four state-of-the-art GE 250MW variable-speed reversible turbines and four 280MVA variable-speed motor generators.

The system added 1,000MW of generation to Switzerland’s hydro resource; its upper reservoir, Lake Mutt, can store 34 Gigawatt hours (GWh) of power; and the plant consistently provides frequency control and other ancillary services to secure the Swiss grid.

Sensor-checking operation and maintenance

Advances in data science and software engineering have also reduced the cost of ownership and increased the reliability of high-value hydro-power assets. Digitisation delivers real-time monitoring and relatable historical data to plant operators, enabling informed decision making in running and maintaining equipment for maximum availability.

Asset performance management (APM) software can, for example help operators understand the consequences of running machinery outside its normal operating range. This type of operation can be justified by a particular market condition, and APM will allow operators to anticipate some of the costs and impacts of their choices.

With digitisation, hydro operations or all kinds can move towards condition-based maintenance systems.

Instead of carrying out maintenance at prescribed intervals, they use sensor-based data to analyse the condition and changes in condition of their equipment, performing maintenance as needed, well in advance of failure or poor performance. This enables significant reduction in the costs associated with asset maintenance.

Globally, established major hydro-electric companies are now working with GE Renewable Energy towards a condition-based maintenance system. Instead of carrying out maintenance at prescribed intervals, they are embedding performance-monitoring sensors in their machinery, and using sensor-based data to analyse the condition and changes in condition of equipment.

This allows them to perform maintenance as needed, well in advance of failure or poor performance.

Pumped-hydro storage has long been a green-energy powerhouse. Research, development and digital analysis continue to reinvent its capabilities as the dynamic, flexible partner to low-cost wind and solar, and a renewable answer to grid security.

This is the fourth instalment of a five part series on pumped hydro sponsored by GE Australia. You can also find a special edition of the Energy Insiders podcast, focusing on pumped hydro, here.

Print Friendly, PDF & Email