WASHINGTON — A study recently completed by Lockheed Martin Space Systems concluded that it would be feasible and cost-effective to launch GPS 3 satellites two at a time on Atlas 5 rockets, a Lockheed Martin official said May 24.

At the same time, Lockheed Martin is working with the Air Force Space and Missile Systems Center to finalize requirements in the coming months for the second block of GPS 3 spacecraft that may include enhancements such as satellite-to-satellite cross-links, said John Frye, Denver-based Lockheed Martin Space Systems’ GPS 3 capability insertion program manager.

The Air Force tapped Lockheed Martin in May 2008 to build the first two of as many as 12 GPS 3A satellites planned to begin launching in 2014. The service now plans to spend $2.65 billion buying a total of eight spacecraft in this first block before moving on to the GPS 3B increment, which also will be developed by Lockheed Martin, Air Force spokesman Hien Vu said in a May 11 email. The second block would begin launching in 2018.

Earlier generations of GPS spacecraft were launched on Delta 2 rockets, which the Air Force decided to stop using in 2009. The first of a dozen GPS 2F satellites, built by Boeing Space & Intelligence Systems of Seal Beach, Calif., was carried to orbit in 2010 by a Delta 4 rocket. Both the GPS 2F and GPS 3 series are designed to launch on either the Atlas 5 or Delta 4 rockets operated by United Launch Alliance (ULA) of Denver.

The Air Force in October asked Lockheed Martin to investigate dual-manifesting GPS 3 satellites, and that study is nearly complete, Frye said in an interview. The company’s analysis indicates that it would be cost-effective to launch two satellites on an Atlas 5, but not on a Delta 4.

“Based on the data [ULA] gave us, Delta didn’t look like it made sense because it looked like it was as expensive to launch two at a time as it was to launch one at a time,” Frye said. “So we focused on Atlas and we did find that it could launch the 3A satellites pretty easily.”

The Air Force’s current plan calls for launching a single 3,650-kilogram GPS 3 satellite on an Atlas 5 in the so-called 411 configuration, which has a 4-meter faring and one strap-on solid-rocket booster, and a single-engine Centaur upper stage. Launching two satellites at a time would require the use of the larger, more expensive 551 configuration, with a 5-meter faring and five strap-on boosters, Frye said.

Dual launches would require ULA to develop a new payload adapter for the Atlas 5, and the GPS command-and-control system would have to be modified with two separate communications channels so that it could link with both satellites while they are still in close proximity postlaunch, he said.

ULA began its own dual-launch study in May at the behest of the Air Force, company spokeswoman Jessica Rye said in a June 2 email. The eight-month study will assess the possibility of launching two GPS 3 spacecraft on a single Atlas 5 or Delta 4, but if the capability is developed it could likely be used for other payloads as well, she said.

Lockheed Martin is interested in dual manifesting because it would save the Air Force money, which in theory could be used for adding new capabilities to  the GPS 3B increment, Frye said. He acknowledged that any money saved on launch costs would not necessarily be available for the GPS 3 program.

The Air Force drafted an initial set of requirements for the GPS 3B increment during the program’s risk-reduction phase, but with an ever-increasing emphasis on affordability, not all of the planned improvements are likely to be made, Frye said. Lockheed Martin expects the GPS 3B requirements to be finalized in time for a system design review planned for August.

One of the improvements could be directional satellite-to-satellite cross-links that could eventually allow the entire GPS constellation to be operated from a single ground site in the United States, Frye said. The original plan called for each GPS 3B satellite to host four cross-links. A cost-benefit analysis shows that scaling back to two cross-links per satellite may be a good idea, Frye said. Another potential upgrade for the GPS 3B increment is a higher gain antenna.

“We understand the current fiscal climate and we’re very interested in working with the customer to make sure they’re getting to a solution that strikes the right balance,” he said. “We’re re-evaluating performance on a number of lines. We’re poking at some of those original requirements, and in some cases we’re finding there’s a soft underbelly. They seemed to make sense at the time, but in light of the analysis we’ve done, we’re offering alternative levels of performance that are a bit lower [and] probably good enough.”