In the vast Gobi Desert near Dunhuang in northwest China’s Gansu Province stands a striking example of what modern renewable energy can achieve: a 100 MW molten-salt tower concentrated solar power (CSP) plant. Surrounded by roughly 12,000 large, computer-controlled mirrors (heliostats), this facility does something conventional solar panels cannot—it generates clean electricity not only during daylight hours but well into the night.
The system works by focusing sunlight onto a central receiver atop a towering structure about 260 meters high. The intense heat raises molten salt to temperatures exceeding 560°C. That hot salt is stored in insulated tanks and later used to produce steam that drives turbines. With thermal storage capacity supporting many hours of generation after sunset (commonly cited around 11 hours under design conditions), the plant can deliver steady, dispatchable power around the clock. It covers several square kilometers, produces on the order of 390 GWh of electricity annually, consumption of approximately 100,000 households, and has been operating commercially since late 2018.
Reliable Energy Production in a Variable World
Traditional photovoltaic solar is intermittent—output drops to zero when the sun sets or clouds gather. Battery storage can help, but large-scale thermal storage in molten salt offers a proven, long-duration alternative that pairs well with existing steam-turbine technology. Facilities like Dunhuang turn solar energy into a more baseload-capable resource. This reliability is critical as electricity demand grows for homes, industry, electric vehicles, data centers, and electrification of heating and transport. By providing predictable power day and night, such plants reduce the need for fossil-fuel backup generation and help stabilize grids that integrate increasing shares of variable renewables.
The technology also demonstrates scalable innovation. Similar molten-salt CSP projects are advancing elsewhere, showing that sunny, arid regions can host large, high-capacity renewable plants that complement wind and photovoltaic farms. In an era of rising energy needs and the push for energy security, dispatchable clean power is not optional—it is essential.
Clear Benefits for Environmental Health
Every megawatt-hour generated by this plant displaces fossil-fuel generation and its associated emissions. Designed annual output of roughly 390 GWh equates to avoiding approximately 350,000 tons of carbon dioxide each year—comparable to taking tens of thousands of cars off the road. Beyond CO₂, it reduces other air pollutants linked to coal and gas plants, improving local and regional air quality and public health.
Because the energy carrier is heat stored in salt rather than continuously burning fuel, the plant operates with near-zero operational emissions once built. Land use in desert areas minimizes competition with agriculture or dense ecosystems, and the closed-loop nature of the molten-salt system limits water consumption compared with some conventional thermal plants. Over its lifetime, the facility contributes to lower greenhouse-gas concentrations, helping mitigate climate risks such as extreme weather, rising temperatures, and ecosystem stress.
A Practical Path Forward
The Dunhuang plant is more than a technological showcase. It illustrates that we already possess workable solutions for producing substantial amounts of clean, reliable electricity while protecting the environment. Scaling similar facilities—alongside efficiency improvements, other renewables, and smart grids—can accelerate the shift away from high-emission sources. The result is energy systems that support economic activity and human well-being without the cumulative damage of fossil-fuel pollution and carbon emissions.
In short, plants like this one matter because they prove renewable energy can be both abundant and dependable. They deliver power when people need it and simultaneously reduce the environmental burden of electricity generation. Expanding and refining such technologies is one of the most concrete steps available toward cleaner air, a more stable climate, and secure energy supplies for the decades ahead.
Overall Assessment
The text is technically well grounded but overly promotional. The Dunhuang plant is real, relevant, and demonstrates that solar thermal energy can be stored and dispatched after sunset. However, the text conflates:
Nominal capacity with actual generation;
Nighttime generation with guaranteed 24-hour supply;
Operational emissions with life-cycle emissions;
Thermal storage with the absence of storage limitations;
Promotional estimates of avoided CO₂ with directly measured values.
My assessment would be: essentially true, but with several important caveats and some misleading claims.
Fact-Checking the Main Figures
| Claim | Assessment | Correction or caveat |
|---|---|---|
| 100 MW solar tower plant | Correct | This is the Shouhang Dunhuang Phase II, a 100 MW electrical CSP facility. |
| Approximately 12,000 heliostats | Correct | The technical database lists 12,121 heliostats, each with an area of approximately 116 m². |
| Tower approximately 260 meters high | Correct | The reported height is 263 meters. |
| Salt heated to more than 560°C | Correct | The nominal temperature rises from approximately 290°C at the receiver inlet to 565°C at the outlet. |
| 11 hours of storage | Correct | The plant was designed for 11 equivalent hours of operation at nominal output. |
| Electricity generation at night | Correct | The stored heat can produce steam and electricity after sunset. |
| Energy “day and night” | Partly correct | This is possible under favorable conditions, but it does not mean guaranteed continuous generation every day of the year. |
| 390 GWh annually | Projected/promotional figure | Some sources cite 390 GWh, while the SolarPACES/NLR database records an even higher original expectation of 483 GWh per year. Actual production remained below these values in several years. |
| Commercial operation since 2018 | Almost correct | The plant was connected to the grid in December 2018 but reached full-load operation only in June 2019. |
| Covers several square kilometers | Correct | The reported total site area is approximately 8 km², with 1.4 million m² of reflective surface. |
| Supplies 100,000 households | Plausible but illustrative | This depends heavily on the assumed household electricity consumption and does not mean that specific homes are supplied directly by the plant. |
| Avoids 350,000 tonnes of CO₂ per year | Conditional estimate | This assumes generation close to 390 GWh and the replacement of highly carbon-intensive coal-fired electricity. It is neither a fixed amount nor a directly measured figure. |
| Closed-loop salt system reduces water consumption | Incorrectly formulated | The closed loop reduces salt replacement, but water consumption depends mainly on cooling the steam cycle and cleaning the mirrors. Dunhuang uses dry cooling, which is what substantially reduces water use. |
The main technical parameters can be verified in the international SolarPACES/NLR database. The plant’s construction and operating history are also described in a technical paper written by people involved in the project.
What “11 Hours of Storage” Really Means
The headline “green energy day and night, without batteries” contains an element of truth but may lead readers to an exaggerated conclusion.
The plant does not use large electrochemical batteries to store its main energy supply. Instead, it stores sensible heat in two tanks:
The relatively “cold” salt, at approximately 290°C, is pumped to the receiver.
Concentrated solar radiation heats it to approximately 565°C.
The hot salt is transferred to the storage tank.
When electricity is required, its heat is used to produce steam.
The steam drives a conventional turbine.
The cooled salt returns to the first tank.
The 11-hour figure means, approximately, that a fully charged thermal store can sustain the turbine at its nominal output for that period. It does not mean that the plant can produce 100 MW during daylight and then another 100 MW for 11 hours without constraints. Part of the solar radiation captured during the day must be used to charge the storage system.
There are also:
Optical losses from the mirrors;
Thermal losses from the receiver, pipes, and tanks;
Losses in the steam-turbine cycle;
Internal electricity consumption by pumps, tracking systems, controls, and heating equipment;
Maintenance periods;
Reduced production due to clouds, dust, and low solar irradiance;
A need to prevent the molten salt from cooling and solidifying.
A more accurate formulation would therefore be:
“The plant can shift a substantial portion of solar generation into the nighttime and, under suitable conditions, maintain continuous generation for extended periods without relying on large electrochemical batteries.”
The Problem with the 390 GWh Figure
A 100 MW plant generating 390 GWh per year would have a capacity factor of:
[
\frac{390{,}000\ \text{MWh}}{100\ \text{MW}\times8{,}760\ \text{hours}} \approx 44.5%
]
This is technically possible for a CSP plant with thermal storage, but it should not automatically be presented as actual production.
There is an important inconsistency among the sources:
Promotional material cites approximately 390 GWh per year;
The SolarPACES/NLR database records 483 GWh per year as the expected generation;
Operational data reported by the industry indicate approximately 200 GWh in 2021 and around 235 GWh in a later year, showing that the learning curve and operating problems were significant.
The text should therefore say “projected annual generation,” rather than simply stating that the plant “produces 390 GWh.” This distinction is especially important because pioneering CSP projects frequently take several years to achieve stable performance.
Is It Really Baseload Power?
“Baseload-capable” is not the best expression. The plant is more accurately described as dispatchable within the limits of its thermal reserve and accumulated solar input.
It can:
Generate electricity during the early evening, when demand often remains high;
Reduce steep output declines associated with sunset;
Shift energy between different hours;
Provide rotational inertia and other services associated with a steam turbine;
Complement photovoltaic and wind generation.
However, it is not equivalent to a permanently firm source such as a hydroelectric plant with a large reservoir, a nuclear power station, or a thermal plant with fuel readily available. After several days of insufficient direct sunlight, the stored thermal energy will be depleted.
Furthermore, CSP primarily requires direct solar radiation. Diffuse light on cloudy days—which can still generate some electricity in photovoltaic panels—cannot be efficiently concentrated by heliostats.
Avoided Emissions: A Possible but Non-Universal Figure
Dividing the estimated 350,000 tonnes of avoided CO₂ by 390 GWh gives approximately:
[
897\ \text{kg of CO₂ per MWh}
]
This value is compatible with replacing relatively carbon-intensive coal-fired generation. However, the amount of emissions actually avoided depends on the marginal source operating on the grid:
If the plant displaces coal, the reduction can be substantial;
If it displaces natural gas, the reduction will be smaller;
If the grid already receives large volumes of renewable electricity or faces transmission constraints, the reduction may be smaller still;
If another renewable source is curtailed instead, the marginal benefit may be limited.
For this reason, the statement that “every MWh displaces fossil-fuel generation” is too categorical. A more accurate statement would be: “Every MWh may displace fossil-fuel generation, depending on grid conditions and operation.”
Nor are emissions absolutely zero. Construction of the tower, production of concrete and steel, manufacture of the mirrors, production of the salts and turbine, maintenance, and replacement of equipment all generate emissions. Life-cycle studies place CSP within a low-emissions range comparable to other low-carbon sources, but not at zero. Estimates by the US Department of Energy put CSP systems at approximately 24–28 g of CO₂ equivalent per kWh, far below coal but above zero (DOE SunShot Vision Study).
It would also be more accurate to say that the plant avoids additional emissions, rather than necessarily “reducing atmospheric greenhouse-gas concentrations.” Reducing concentrations that have already accumulated would require net-negative emissions.
Water Use and Environmental Impacts
The text is correct in suggesting that Dunhuang consumes less water than many conventional thermal power plants, but it incorrectly attributes this advantage to the closed-loop salt system.
The plant uses dry cooling, as confirmed by the project’s technical documentation. This drastically reduces the amount of water needed to condense the steam. Studies estimate that dry cooling can reduce CSP water consumption by approximately 71–78% compared with wet cooling.
Even so, water is still needed for:
Cleaning more than 12,000 mirrors;
Producing and replenishing water for the steam cycle;
Services and maintenance;
Dust control during certain operations.
Dry cooling also comes at a price: it costs more and may reduce turbine efficiency on very hot days.
The desert should not be treated as an environmentally empty space. A facility covering approximately 8 km² may cause:
Fragmentation of arid habitats;
Changes to soil and drainage;
Effects on birds and insects near the receiver;
Glare;
The need for roads and transmission lines;
Material consumption and disturbance during construction.
These impacts are probably far smaller than the climatic and atmospheric damage caused by coal, but they still need to be included in the comparison.
The Omitted Economic Limitation
The greatest omission in the text is cost. CSP with storage is valuable, but it normally costs more than photovoltaic or wind power without storage. The SolarPACES database reports the following figures for Dunhuang:
An investment of approximately US$458 million in 2020 values;
A specific cost of around US$4,582 per kW;
An original remuneration tariff equivalent to approximately US$0.17 per kWh;
Support through China’s demonstration program.
The International Energy Agency recognizes the systemic value of CSP but notes that high costs, technological risks, financing difficulties, and construction complexity have limited its expansion. However, the fair comparison is not simply “CSP versus photovoltaic panels.” It should be:
CSP with 8–15 hours of thermal storage versus photovoltaics plus batteries, grid reinforcement, backup generation, and grid-stability services.
For short-duration storage, the photovoltaic-battery combination is frequently more economical. For prolonged nighttime storage, CSP may remain competitive in regions with exceptionally strong direct solar radiation. The IEA highlights both its higher cost and its value as a dispatchable resource.
Critical Conclusion
The Dunhuang plant is an important achievement, but it does not prove that solar thermal energy is a universal solution or that it can provide uninterrupted electricity without constraints. What it demonstrates is something more precise and still highly valuable:
In regions with strong direct solar radiation, a CSP plant can store large amounts of energy as heat at a potentially competitive cost for long-duration storage, shift generation from daytime to nighttime, and complement photovoltaic, wind, hydroelectric, nuclear, and other firm energy sources.
The text’s central message should therefore be preserved, but “abundant and dependable 24-hour energy” should be replaced with “partly dispatchable solar energy with long-duration thermal storage.” This wording is less spectacular but scientifically more defensible.
No comments:
Post a Comment