ROBOT.LK

Astronomy produces a lot of carbon emissions, but it could be one of the greenest sciences if observatories harness their solar and wind resources

IF YOU were to draw up a list of the most pressing issues in science, it's unlikely that astronomy's carbon footprint would be on it. If it were, it would probably end up somewhere between effective male birth control and how to fold headphones to stop their wires getting tangled in your pocket.

Ueli Weilenmann, deputy director of La Silla Paranal Observatory in Chile, would disagree with that assessment. Recently, while grappling with the costs of running the place, he was shocked to discover the scale of the observatory's carbon emissions (see diagram). A bit of further digging revealed that the problem is not limited to Paranal: many other observatories exude more greenhouse gas than their size betrays.

This shouldn't be the case. By dint of their location, most observatories enjoy access to clean energy sources, but for various reasons they have been unable to exploit them. Now observatories all over the world are looking beyond obvious solutions, enlisting ingenious workarounds in their quest to go green. The possibilities for doing so run from the inspired to the mundane to the highly speculative. The potential carbon cutbacks won't save the world, but the people running these experiments are determined to prove that big science can be clean too.

The bigger telescopes get, the further they can peer into our universe, and the better the resulting images. But the barbed spiral galaxies and weather on distant exoplanets that have been captured by Paranal's Very Large Telescope (VLT) come at a high cost. Astronomy is an energy-intensive endeavour. "We are in a very isolated place and everything we do here has an associated energy cost," Weilenmann says. Paranal is so remote that even the water needs to be trucked in, not to mention food, staff and fuel.

The lion's share of the energy use, however, comes from running an instrument like the VLT and cooling its sensitive electronic equipment. Every day it sucks up 27 megawatt-hours of energy, or nearly 10 gigawatt-hours per year – the annual consumption of 1000 US households.

But unlike those homes, Paranal is too far from the national grid to connect, so it must produce its own power. It does this using generators that burn butane. Fuel prices are volatile, and with observatories hardly swimming in cash, Weilenmann was investigating Paranal's energy use to try to keep expenses under control. It was then that he discovered its carbon footprint, 22,000 tonnes a year, equivalent to 46 tonnes of carbon dioxide for every peer-reviewed scientific paper produced there. It's equivalent to the emissions of a small town.

In a world where the energy budget of a data centre can rival that of a medium-sized city, those numbers won't raise many eyebrows, but for Weilenmann it was a matter of principle: the problem should not have existed in the first place. After all, the ideal locations for observatories happen to be green-energy sweet spots. "We never faced a situation where there was no sun and no wind for more than a day," says Rolf Chini of Ruhr University Bochum in Germany. Chini runs the observatory at Cerro Murphy which, like Paranal, sits on a peak in the Atacama desert, with 320 cloudless days a year on average and buffeted by strong winds.

There's just one hitch: neither wind nor solar are straightforward energy sources for observatories. Solar panels cannot power a telescope at night. High winds are similarly problematic – great for power generation, but terrible for stargazing. On the windiest nights, observatories need to keep their doors closed. But even on mellower nights, turbines whip up turbulence that can interfere with stargazing.

And yet, Chini's observatory had somehow managed to do it. Billing itself as the world's only 100-per-cent green observatory, it obtains its energy exclusively from 100 solar panels and three wind turbines.

When he heard of this, Weilenmann set out to learn lessons for Paranal. His hopes were quickly dashed, however, when it became clear that these green credentials were all down to the observatory's small size. The three buildings are the size of small garages, with turbines that, at 5 metres across, are too small to create turbulence. Solar and wind energy produce liquid nitrogen stored to cool the telescopes' sensitive electronic equipment at night, and any surplus energy is stored in banks of batteries to power the telescope at night. These batteries, which hold at most 2.3 kilowatt-hours, can supply just enough power to run a toaster and an Xbox at the same time. Paranal's telescope would need about 500 times more juice, far beyond the capability of any off-the-shelf battery storage system. Rigging up their own battery banks, Weilenmann says, would require at least 1200 batteries, "a €600,000 investment" that would need to be replaced within 10 years. Not exactly an improvement over butane.

Nonetheless, having calculated that switching to renewables could reduce Paranal's carbon footprint by about 43 per cent, he was determined to find a way to do it. Other observatories had worked out their own solutions, some of which exploit more localised sources of green energy. Leading this effort is the Square Kilometre Array (SKA), a colossal 3000-antenna radio observatory that will be completed in South Africa and Australia in 2024. Already up and running are its highly sensitive "precursor" telescopes, which need to be cooled day and night, and generate large amounts of data to be crunched by supercomputers, which also need cooling. At SKA's supercomputing centre in Perth, Australia, engineers use geothermal pumps to retrieve groundwater at 22 °C from an aquifer 140 metres below, pumping it through a heat exchanger to cool their systems and reinjecting it into the aquifer 11 °C hotter. This trick doesn't just save energy. "If cooling towers were used instead, the evaporative water loss would equal 15 Olympic swimming pools a year," says Jerry Skinner, SKA's public affairs officer.

However, geothermal energy is not available in most places, and the world's most ubiquitously available clean energy source is solar, so other initiatives are under way to work around its idiosyncrasies. Solar panels can be used not just for day-to-day needs but also to pre-cool liquid nitrogen for the telescopes to use at night. Even if a photovoltaic system could partially supply Paranal's energy during the day, says Weilenmann, it would cut the observatory's yearly fuel use by 18 per cent. Thinking along similar lines, staff at the South African Astronomical Observatory – which houses the Southern African Large Telescope – wanted to build a solar farm, but the parent organisation would not provide funding. So they improvised, rigging up "solar geysers" that concentrate sunlight to heat water in pipes. Thanks to this and other small measures, the observatory's energy consumption has dropped. "We have managed to reduce the kWh consumption by 11.9 per cent per month," says Keith Browne, a member of the operations team.

Even if observatories had money for solar farms, however, there would still be another issue. To meet Paranal's peak daytime power demand – 600 kilowatts – using standard solar panels, Weilenmann says, "we would have to install 2400 of them". That's a lot of panels.

Intriguingly, a partial solution might be found in the light-harvesting abilities of telescopes themselves. While developing the giant honeycombed mirrors to be used in the future Giant Magellan Telescope, Roger Angel, an astronomer at the University of Arizona in Tucson realised the mirrors could be used to focus sunlight rather than starlight. He founded a company, Rehnu, to develop the concept: several large squarish parabolic mirrors attached to a lightweight steel frame, concentrating sunlight onto a matrix of small photovoltaic cells. His prototypes have demonstrated an efficiency of about 40 per cent, Angel claims – more than double that of traditional solar power systems. If he's right, observatories could build modest solar farms that would nonetheless put an appreciable dent in their carbon footprint.

After a long and largely frustrating hunt for solutions, Weilenmann's salvation came from the last place he expected – the grid. While he had always known the grid offered the best hope for greening Paranal, the case for doing so had long been thought closed: there was just no way to connect Paranal to Chile's distant, privately operated grid. But then the green light was given to build the European Extremely Large Telescope, scheduled for completion within a decade at Cerro Armazones, 22 kilometres away. "There is no doubt E-ELT will get a grid connection," Weilenmann says. Its power line will cut through the Atacama and right through Paranal – connecting it and finally making it financially realistic for Paranal to invest in solar and wind, generating clean energy by day, selling the surplus and benefiting from feed-in tariffs. "With a grid connection we could produce and supply any amount of energy," Weilenmann says. "The excess would go back to the grid during the day, and during the night we would recover it from the grid again."

It would only be fitting for an observatory to be powered by its own star. Now, about those headphone wires...

This article appeared in print under the headline "Green sky thinking"

Moon units

Solar power is the best option for most observatories: clean, plentiful, cheap and off the grid. In practice, however, efforts to use this intermittent source have been stymied by many obstacles (see main feature).

One way to make solar power a 24-hour resource involves capturing the sun's weak reflection off the moon. This idea won't see the market any time soon, but plenty of companies are working on it. Rawlemon in Barcelona, Spain, claims that its weatherproof glass globes can collect sunlight by day and moonlight by night. A 500-millimetre sphere can track the moon, concentrate its thin stream of reflected photons up to 20,000 times onto a standard solar cell, and produce enough electricity to operate an LED, says Rawlemon's André Broessel.

A more exotic proposal, floated by the Shimizu Corporation in Japan, is to turn the moon itself into a solar panel. The company's Luna Ring concept involves putting a "solar belt" around the moon. This 11,000-kilometre-long, mirrored structure will generate electricity from sunlight and beam it back to Earth with lasers.

In light of recent interest in rehabilitating a long-dismissed plan to beam solar power from space, such ideas might not be so far out.