WASHINGTON — NASA is investigating a detector problem common to three primary instruments on the James Webb Space Telescope (JWST) that could get worse with time as a planned 2014 launch date for the overbudget flagship-class astronomy mission slips further into the future.

NASA chartered a Failure Review Board Feb. 7 to assess the problem, which affects detector arrays made by Teledyne Imaging Sensors of Camarillo, Calif., for JWST’s Near Infrared Camera (NIRCam), Near Infrared Spectrograph (NIRSpec) and Fine Guidance Sensor-Tunable Filter Imager.

Marcia Rieke, principal investigator for the NIRCam, the telescope’s primary imager, said tests conducted over the past two years on five long-wavelength detector arrays revealed so-called hot pixels that degrade the quality of images JWST is designed to capture. That is disappointing news for a telescope expected to cost more than $5 billion — and possibly much more, according to the findings of an independent review conducted last year. But the real problem, Rieke said, is that the pixel degradation appears to be growing worse with time.

“If this were all that might happen, this would not be an issue,” she wrote in a Feb. 21 e-mail. “But if the array continues to degrade with time until launch, too many pixels would be out of spec.”

Two years ago when Rieke’s NIRCam team tested the detector arrays at the University of Arizona in Tucson, the hot-pixel problem was evident, yet remained well within the limits set by JWST’s stringent performance requirements. But in December, as the team was readying a lone detector for integration into the NIRCam instrument at Lockheed Martin Advanced Technology Center in Palo Alto, Calif., final performance test data indicated the problem had worsened.

“Much to our horror, the test revealed that there are now about 2 [percent] of the pixels that are out of spec rather than the initial 0.5 [percent],” Rieke told Space News, adding that a subsequent test in January showed four out of five of the NIRCam’s long-wavelength detector arrays were affected.

But Rieke said the problem is not isolated to the NIRCam detector arrays. Teledyne-built detectors designed for the NIRSpec instrument, developed by Astrium GmbH of Germany, and the Fine Guidance Sensor-Tunable Filter Imager, in development by the Canadian Space Agency, are affected as well, she said.

“The problem we are having is that just sitting on the shelf under controlled conditions, some initially good pixels went bad this way,” she said, explaining that the technology used to build the infrared detector arrays involves the same solid-state physics as pixels used to capture images in digital cameras, in which hot pixels fail to yield data. “No one would want to buy a digital camera that developed bad pixels just sitting around!”

Rieke said the Failure Review Board, which is expected to complete work in April, hopes to identify exactly what is causing the degradation and whether it might be reversible.

“It is certainly a serious problem because we do not understand it today,” she said.

Led by NASA’s Goddard Space Flight Center in Greenbelt, Md., JWST is an infrared telescope with a 6.5-meter foldable mirror and a deployable sunshield the size of a tennis court. Northrop Grumman Aerospace Systems of Redondo Beach, Calif., is prime contractor.

An Ariane 5 rocket provided by the European Space Agency is slated to launch the observatory to the second Lagrange point — a gravitationally stable spot 1.5 million kilometers from Earth — in June 2014. However, an independent review team convened last year found the estimated $5 billion telescope to be at least $1.5 billion over budget and 15 months behind schedule. The panel urged NASA to furnish $500 million in additional funding in both 2011 and 2012 or risk further slips to the launch delay.

Rick Howard, JWST program director, said Feb. 14 that the additional money is not likely to materialize, and that JWST’s launch date could slip well into 2016 as a result. He said continued launch delays could pose a problem for JWST’s instruments, including the detector arrays.

“As you get further and further out with [the launch date], it really raises questions about how far down the [integration and test] process you go for the instruments … and how long you have to store all that before you actually launch,” he told the NASA Advisory Council’s astrophysics subcommittee during a Feb. 16 public meeting here. “And that just makes everybody even more nervous about this problem than anything else.”

Howard said NASA has considered scrapping the existing detectors and buying new ones prior to JWST’s launch.

“We have had discussions about why not just go off and have another lot of detectors made up, and this is probably what we may do. But we need to find out what the root cause is rather than just replicate the problem that we now have two years from now,” he said. “But that is one of the options.”

Rieke said having to buy new detectors just prior to launch would be a worst-case solution to the pixel degradation plaguing JWST’s sensors.

“I have a lot of hope that we will find a simpler solution,” she said.