The head of the Veterans Administration said yesterday that the cost of compensating Vietnam veterans for harm done them by the Agent Orange herbicide would be hundreds of millions of dollars per year for many years.

At the same time, the UCLA doctor planning the VA's study of Agent Orange in Vietnam, said that it may never be possible to determine exactly who was exposed to it and how much exposure they received.

Other witnesses before a hearing at the Senate Veterans' Affairs Committee said the VA study, already two years in the works without even a study protocol set, was in such difficulty that the VA should consider scrapping current plans and starting over.

When Robert P. Nimmo, the new head of the VA, was asked about the consequences of identifying soldiers who may have been harmed by Agent Orange and the cost of compensation should the harm be widespread, he responded: "We would be looking at hundreds of millions of dollars per year, going into the middle of the next century."

Senator Arlen Specter (R-Pa.) later suggested that the prospect of such a massive compensation program could be one factor in the VA's delay on Agent Orange questions.

Nimmo also acknowledged that the study of Agent Orange has just rejected the study protocols worked out by Dr. Gary Spivey and his colleagues at UCLA. Witnesses from the agencies that reviewed the plans found them "completely inadequate." Vernon Houk of the Center For Disease Control said the proposal "had such insufficient information that we did not indeed even classify this as a protocol."

The trouble with the protocols, which were already late, was that the VA and UCLA had assumed they could have access to military records to carry out the study. But they found that the only good records were of where the herbicides were sprayed.

The records of where the soldiers, or at least their units were in action in Vietnam may be impossible to use. They occupy 40,000 shelf-feet, or almost eight miles of records, are not indexed, and must be sorted by hand.

Even a lengthy hand sorting, said Roger Detels, dean of the UCLA School of Public Health, "might not give you what you need" to determine how much exposure soldiers or their units actually got.

"The VA is simply not competent to do this study," concluded Ronald Simon of the National Veterans Law Center.