WASHINGTON — Lockheed Martin Space Systems is shrinking a team of 140 down to 25 now that NASA has canceled work on the Advanced Stirling Radioisotope Generator (ASRG), a next-generation nuclear battery for planetary missions that cannot rely on solar power.

It will cost about $2 million and the effort of 10 employees to close out the program, Bob LeRoy, Lockheed’s director of East Coast Operations, said in a Jan. 14 phone interview. Another 15 employees, at a cost in the “single-digit millions” range, will continue work through mid-2014 on ASRG hardware NASA wants completed, LeRoy said. That work will focus on the device’s controller, which is required to change the alternating current the ASRG produces into direct current compatible with a spacecraft bus.

NASA had been spending about $55 million a year on ASRG development and will save roughly $170 million during the next three years by canceling the project, Jim Green, director of NASA’s Planetary Science Division, told the agency-chartered Outer Planets Advisory Group at a Jan. 14 meeting in Tucson, Ariz.

Because ASRGs are fueled with plutonium, the Department of Energy managed Lockheed Martin’s ASRG development contract, which was awarded in 2008. However, NASA paid for all the work. Since 2008, NASA has spent $272 million on ASRG, Len Dudzinski, program executive for radioisotope power systems at NASA headquarters here, wrote in a Jan. 15 email to SpaceNews.

Most of that went toward Lockheed’s contract, which was worth about $304 million when a stop-work order arrived in December. The contract includes a $45 million extension and a $25 million cost overrun, both of which occurred in 2012, LeRoy said. For its money, NASA got a final design for the ASRG, plus a partially completed qualification unit, which will be shipped to NASA’s Glenn Research Center near Cleveland later this year.

“NASA is planning to continue technology maturation of Stirling power conversion technology for space applications,” Dudzinski said. “The Stirling convertor laboratory at Glenn will receive this hardware and integrate it into its existing testing program.”

The ASRG controller is being built in Denver. Hardware for the qualification unit was being developed at Lockheed’s Valley Forge, Pa., facility, which is not part of the Newton, Pa., facility the company is closing as part of the consolidation it announced late last year.

To enable a wider range of destinations for smaller spacecraft, the Lockheed-designed ASRG was to be lighter and more efficient than the current-generation Multi-Mission Radioisotope Thermoelectric Generator (MMRTG), which is built by AerojetRocketdyne. The Mars Science Laboratory uses an MMRTG, as will its planned successor, the Mars 2020 sample-caching rover. The flagship Cassini Saturn orbiter that launched in 1997 and the New Horizons probe that will reach Pluto in July 2015 use the previous-generation nuclear battery, a Lockheed product called the General Purpose Heat Source Radioisotope Thermoelectric Generator (GPHS-RTG).

Stirling engines are not new, although NASA has never used them on radioisotope power systems for spacecraft. Back in 2000, when Lockheed bid on and won a study contract to examine the viability of a Stirling generator for a next-generation radioisotope power supply, “we thought that was the future,” LeRoy told SpaceNews. NASA thought so too.

In the agency’s last competition for the Discovery program, a class of small, scientist-led planetary missions where costs are capped at roughly $425 million, NASA offered an ASRG as government-furnished equipment. Two of the three finalists, a comet-hopping probe lander and a spacecraft designed to sail the hydrocarbon lakes of Titan, expressed interest in the unit, which was to be provided in time for a 2016 launch. However, when NASA made its final selection in September 2012, it picked a solar-powered Mars lander called InSight, which Lockheed will build and operate.

An ASRG could produce about 140 watts of power using slightly less than a kilogram of plutonium-238. That is less than a quarter of the plutonium the heavier MMRTG uses to produce about the same wattage.

But with NASA’s planetary science budget shrinking, even as the division takes on new responsibilities and struggles to start new missions while maintaining flagship spacecraft with no planned successors, something had to give. It gave back in November, when NASA decided that MMRTGs were efficient enough, given the agency’s ongoing efforts to rebuild the U.S. stockpile of plutonium-238.

“With the increase in plutonium and our ability to continue on and build or create new plutonium-238, it’s not like we closed the door on use of radioisotope power systems completely,” Green said in Tucson.

Even as NASA was paying for ASRG development, it was forking over roughly $50 million a year to support the Energy Department’s restart of plutonium-238 production. About 30 aging kilograms of that isotope remain in U.S. stockpiles, and the production restart NASA is funding would add about 1 kilogram of the substance a year, beginning in 2021.

When it came time to find a bill-payer for plutonium production, stopping the ASRG program made more sense than taking the funds out of the Discovery program or Planetary Science research grants, Green said.

MMRTG, or Nothing

Both ASRGs and MMRTGs use heat from decaying atoms of plutonium-238 to create electricity. While the lighter, more efficient ASRG could potentially have opened hard-to-reach parts of the solar system to cheaper missions, all radioisotope power systems come parcelled with increased risks that could drive up missions costs.

For example, it cost around $65 million just to study whether a launch mishap would have exposed the public to plutonium-238 during Cassini’s 1997 launch to the Saturn system, Ralph McNutt, a planetary scientist based at the Applied Physics Laboratory in Laurel, Md., told the Outer Planets Assessment Group.

So, for now, planetary science missions that need radioisotope power at all will have to stick with the MMRTG.

“There are seven of these units that could be fueled,” McNutt said

Of the seven MMRTGs, one — plus a flight spare — belongs to the Mars 2020 rover, which will be based on the design for the Mars Science Laboratory and launch to the red planet in 2020 to cache and store a sample of the martian surface. The estimated cost of this mission is $1.5 billion. Of the remaining MMRTGs, four are being reserved “for a large mission later on,” Green said.

Meanwhile, the project team for a proposed Europa mission called Clipper is now working to reduce the number of MMRTGs needed for its mission concept to four from five.

Clipper would notionally launch in November 2021 aboard a United Launch Alliance Atlas 5 551, Brian Cooke, Europa Clipper’s project system engineer, told the Outer Planets Advisory Group Jan. 14.

NASA still has made no formal commitment to the Clipper mission, which is popular enough in Congress that lawmakers approved $80 million worth of study money for the concept in the Consolidated Appropriations Act for 2014 (H.R. 3547). That is about $10 million more than the mission team had to work with in 2013, when federal budgets were fully sequestered in accordance with the Budget Control Act of 2011.

Barry Goldstein, the Europa Clipper project manager, told the Outer Planets Assessment Group he hoped to see enough funding in the White House’s 2015 budget request to formally begin the mission in the next federal budget year, which begins Oct. 31. However, the Clipper team will not have a solid estimate for the mission’s baseline cost until the end of 2014, when Clipper’s mission concept review will take place, Goldstein said.

Dan Leone is a SpaceNews staff writer, covering NASA, NOAA and a growing number of entrepreneurial space companies. He earned a bachelor’s degree in public communications from the American University in Washington.