IN ADDITION to the superconducting supercollider at $10 billion, the human genome project at $2 billion and Space Station Freedom at $38 billion, other large-scale research projects planned for the coming decade include an Earth-observing satellite system to monitor global climate change, projected to cost $17 billion. Fusion energy research will cost $5 billion over the decade, with little hope of practical returns before 2050. And there are many smaller projects, including a gravity wave observatory for $47 million, a $60-million high magnetic field laboratory and an unmanned spacecraft to rendezvous with a comet at $158 million -- small compared to supercolliders or space stations, but nevertheless huge on the scale of the individual investigator.

Most of these projects are just getting underway. If they proceed as planned, they will reach their maximum rate of expenditure four or five years down the road, producing a devastating annual budget crunch while the nation is still trying to dig out from under the rubble of the collapsed S & Ls. Congress, however, has been consistently misled about the eventual cost of these mega-science projects.

The Hubble space telescope was initially estimated at $300 million but ended up costing $2.1 billion -- and it still gives fuzzy images. The estimate for Space Station Freedom, which began at $8 billion, has soared to $38 billion -- and counting; it seems to be the only thing in NASA that consistently goes up.

The first number Congress heard for the superconducting supercollider was $4.4 billion, but when a budget request actually came down from the president, it was for $5.9 billion. The earlier figure, it was explained, was not corrected for inflation. The $5.9 billion estimate was challenged by the Congressional Budget Office, which pointed to previous cost overruns on accelerators and suggested that a 46 percent overrun would be typical. The Energy Department ridiculed the report; its director of energy research indignantly declared that SSC technology is so well understood that the cost can be calculated with complete confidence.

But one year later, the SSC design group reported that a redesign was needed to ensure success; the price tag went up another $2 billion. Now it seems the SSC will need additional detectors once it is up and running another $2 billion. Congress seems to regard all this as normal.