PARIS — Orbital Sciences Corp., which has built its success on providing small satellites and launch vehicles to commercial and government customers, is considering an expansion of its rocket business to be able to carry much larger satellites into orbit, Orbital Chief Executive David W. Thompson said.

Such a move — like Orbital Sciences’ decision to invest in small commercial telecommunications satellites several years ago — would buck conventional industry wisdom, especially among publicly traded companies.

In both the United States and Europe, satellite launching is viewed as a low-margin business at best and investment in new vehicles can be justified only by substantial guaranteed government backing, the presence of a rich entrepreneur — or both.

In an April 19 conference call with investors, Thompson said the Dulles, Va.-based Orbital Sciences nonetheless sees a market opening to enlarge its rocket portfolio beyond the aircraft-launched Pegasus, the ground-launched Taurus and the Minotaur, which is based largely on excess strategic missile hardware.

“There is a much bigger adjacent opportunity for us to move up from 1-2 metric tons of satellite capacity to, say, 4-5 metric tons,” Thompson said. “We’re looking very seriously now at potentially doing that in a way that would make sense to customers who have relied on an existing product in that category that seems to be fading away — and that also would be a good business opportunity.”

Launch vehicles in 2006 accounted for about 40 percent of Orbital Sciences’ total revenue of $790 million, but the three rockets designed to place satellites into space were just 9 percent of the total, behind suborbital interceptor and target vehicles for the U.S. Defense Department and the U.S. Missile Defense Agency.

Orbital Sciences currently competes with the German-Russian Eurockot Launch Services GmbH venture of Bremen, Germany, in the market to launch relatively small government science and Earth observation satellites, mainly for government customers. Eurockot uses a converted Russian ballistic missile.

Orbital’s view of Eurockot was spelled out clearly April 11 in remarks by Garrett E. Pierce, the company’s chief financial officer, in a presentation to investors.

“The question is what the competitive advantage is for Orbital relative to the European-Russian rocket,” Pierce said. “If you want a relatively cheap rocket, but it’s not highly reliable, you’ll buy the Russian-European rocket. If you want a reliable rocket that’s affordable but priced fairly, you’ll buy our rocket.”

Meanwhile, Orbital Sciences’ decision to invest heavily in a line of commercial telecommunications satellites continues to pay off. This product line, featuring the star-2 satellite platform, accounted for one-third of the company’s 2006 revenue.

That performance likely will be repeated or surpassed in 2007. The company expects to deliver five commercial satellites this year and already has booked orders for three new spacecraft — two for Washington-based Intelsat and one for Optus Networks of Australia.

Thompson said a fourth commercial telecommunications satellite order is expected in the next month. Industry officials said Orbital is one of the favored bidders in the competition for an order from Telenor Satellite Broadcasting of Norway. Telenor’s last satellite purchase, in 2004, was from Orbital Sciences.

Satellite-fleet operator SES Global of Luxembourg also is weighing the purchase of new satellites for its SES Americom, SES New Skies and SES Astra divisions.

Thompson said that with three new orders and one option already booked for 2007, Orbital Sciences is on track to win five commercial telecommunications satellite contracts in total this year, a harvest that he said would correspond to about half the expected global market for spacecraft of this class .

The U.S. dollar’s recent weakness against the euro will help Orbital Sciences against its European competition, but the field of players is growing. Astrium Satellites of Europe has joined forces with India’s Antrix Corp. to produce commercial satellites, and OHB System of Germany, backed by European government funding, expects to enter this market starting in 2010.

Small-satellite specialist Surrey Satellite Technology Ltd. of Britain also is targeting the commercial geostationary satellite market.

Orbital Sciences currently produces a telecommunications satellite in about 24 months but expects to be able to reduce that to 21 months as part of a series of manufacturing improvements to be introduced in the next 18 months, Thompson said.

Orbital Sciences in 2006 received $1.2 million in incentive payments from Earth observation satellite data provider GeoEye Inc. for the performance of the Orbital-built OrbView-3 high-resolution imaging satellite. A similar payment was expected this spring.

In March, GeoEye of Dulles reported that OrbView-3’s camera had failed and most likely cannot be recovered. The satellite was launched in June 2003 and had been expected to operate for more than seven years.

Thompson said Orbital expects to receive most of its 2007 incentive payment despite the camera failure insofar as GeoEye “got the economic value of the satellite for roughly 10 [months] of the 12-month period that the current-year incentive would apply to … But it remains to be seen whether that will materialize in the second quarter.”