WASHINGTON — Should Orbital Sciences Corp. convert its Antares rocket into an all-solid-fuel vehicle, the company will have to overhaul its launch infrastructure on Virginia’s eastern shore and figure out how to comply with a NASA launch safety rule written to protect area bystanders from broken glass.

The potential safety show-stopper has to do with a phenomenon known as distant focusing overpressure: shockwaves created by explosions. If a rocket explodes in midflight, the shockwaves could bounce off the ground or even clouds to shatter windows many kilometers from the blast zone.

NASA’s Range Flight Safety Program document therefore requires launch operators to calculate, based on up-to-the-minute weather conditions fed into complex computer models, which areas are at risk for shattered glass.

According to two people familiar with recent discussions between Orbital and NASA, a midflight explosion of a solid-fueled rocket the size of Antares could shatter windows — and thus pose a safety risk — so far from the vehicle’s flight path that the agency would sooner prohibit its launch than undertake the effort required to evacuate buildings in the affected area.

Antares launches from NASA’s Wallops Flight Facility on Wallops Island, Virginia, a range that until recently was used primarily for suborbital sounding rocket missions. The facility is now regularly used for launches of orbit-bound vehicles including Antares, which Orbital uses to launch cargo to the international space station, and Minotaur, a family of rockets assembled by Orbital and based in part on ICBM motors.

The first stage of Antares is powered by two Russian-built NK-33 engines, which are imported and refurbished by Aerojet Rocketdyne of Sacramento, California. Dulles, Virginia-based Orbital is looking at replacement options due to the limited supply of the liquid-fueled engines, which have been out of production for decades.

Among the options under consideration is a solid-fueled motor built by ATK Aerospace and Defense Group of Magna, Utah, which is slated to merge with Orbital under a deal expected to close before the end of the year. Orbital also could opt to restart production of the NK-33, the refurbished variant of which has been rebranded as the AJ-26. The company also could opt to purchase another Russian-made engine such as NPO Energomash’s RD-191, which shares development heritage with the RD-180 that powers United Launch Alliance’s Atlas 5. Alternatively, Orbital might help fund development of a new U.S.-made liquid engine, such as the AR-1 concept Aerojet Rocketdyne has touted in response to increasing appetites among Washington policy makers for a domestic alternative to Russian engines.

Orbital Chief Executive David W. Thompson said recently that a decision on the future Antares first stage is imminent.

In an Aug. 21 email, NASA spokesman Jeremy Eggers acknowledged NASA is exploring tweaks to the models used to calculate the risk posed by distant focusing overpressure. But he said “this is an activity that is beneficial to, but not driven by, Orbital’s current or future launch plans.”

Eggers said Wallops started these exercises in September after another Orbital launch — of NASA’s Lunar Atmosphere and Dust Environment Explorer aboard a solid-fueled Minotaur 5 — was held up because NASA’s overpressure models showed potential shattered-window risk to occupied structures on Wallops Island. The Minotaur 5 is much smaller than the medium-class Antares.

The overpressure issue is not the only hurdle to launching an all-solid-fueled Antares from the state-operated Mid Atlantic Regional Spaceport on Wallops. Orbital also would have to modify one of its two launch pads at the facility to accommodate the vehicle.

Specifically, Pad 0B at the spaceport, which is optimized for Minotaur launches, would have to be enlarged and outfitted with the rigging needed to vertically stack an all-solid Antares, according to Dale Nash, executive director of the Virginia Commercial Space Flight Authority that operates the spaceport.

Alternatively, Orbital could modify the newer Pad 0A, which was built at a cost of $90 million to the state of Virginia and optimized for the current liquid-fueled Antares, Nash said. “We’re willing to work with them either way,” Nash said.

Orbital spokesman Barron Beneski declined to comment for this story.

Increasing the urgency of settling on a long-term solution for the Antares core stage, NASA plans to release a solicitation by Sept. 30 for a follow-on space-station cargo-delivery contract. Orbital plans to bid on the contract, which would require deliveries from 2017 to 2024, the year to which the White House has proposed extending space station operations.

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.