It looks like a Rube Goldberg original looming up there with its huge metal tepee, yellow silo, endless conveyors shooting skyward, spaghetti-like pipes, bright orange graphics and an elaborate name - pyrolysis plant.

Not long ago, the urban prophets were saying the plant could revolutionize the nation's solid waste industry. It would be an environmental dream, turning common, everyday trash - raw garbage, tin cans and bottles - into steam, scrap iron, char and a glassy substance, which could be used to pave streets, all in a short, clean process.

But the plant, originally scheduled to begin operations in August, 1974, has never worked - at least for any extended period of time.

"The possibility of failure never entered anyone's mind," said Francis W. Kuchta, city director of public works. But conveyor belts clog; there has been an explosion or two, and once in the heating process the garbage fused and had to be broken up with jackhammers.

It has left red faces all around. When the plant's private developer, Monsanto Environ-Chem Systems, gave up on the project last month and recommended it be abandoned, an embarassed Mayor William Donald Schaefer angrily called the firm "a bunch of common bastards" who "sold us a bill of goods."

The city didn't abandon the plant. It almost can't. It needs something to get rid of its trash.

But faced with the prospect of shelling out another $12 million to make the pyrolysis plant work, officials are admitting they've learned some hard lessons in their bout with technology - lessons other cities might well heed.

"A city can't afford to experiment. Cities need a sure thing. They don't have money to throw away," said Kuchta. "Solid waste is a sure thing. It isn't going to go away."

It's because of this that the city is giving the plant, which previously never worked more than 18 days at a time, three more 30-day tests. The results, since Monsanto pulled up stakes Feb. 1, has been mixed. But as sanitation director Edward J. Moore said this week, "You don't give up on a $22 million investment."

In hindsight, it was Baltimore's flirtation with an unproven technology that led to its current dilemma. It bought a promising experiment and hoped to expand on that.

A little background is on order here. Back on 1971, Baltimore was looking for a replacement for one of its two old, pollution-belching municipal incinerators, something that could deal with half the city's garbage, or about 1,000 tons a day.

Along came Monsanto and the federal government.

The St. Louis firm had just developed a 35-ton a day solid waste pilot plant. It used a process called pyrolysis, which converted trash into combustible gas by heating it under an extremely high temperature in an oxygen-free kiln. Steam was produced as a byproduct. The process, by all accounts, worked great in the small pilot plant.

About this time, the Environmental Protection Agency (EPA), equipped with new legislation, was fishing around for places to demonstrate new technoligies. It was high on Monsanto, and had money to spend. Needs and money merged.

Enticed by a $6 million EPA grant, Baltimore coughed up $5 million and coaxed another $4 million from state coffers to get the plant started.

The plant began to mushroom, just south of downtown Baltimore, off Russell Street, the main entrance to the city from the Baltimore-Washington Parkway. A graphic artist was hired to add pizzazz to the thing, and early rave reviews appeared in the national press.

But almost immediately the pyroloysis plant developed problems, the original opening date proved overly optimistic. Cost skyrocketed. Anti-pollution divices proved inadequate. Monsanto, which had agreed to a stiff performance bond, shelled out $4 million in late penalties.

And the thing still didn't work. Conveyors broke and clogged. The huge steel barn where garbage entered the plant collected so much carbon monoxide it was dangerous for workmen to stay inside. When trucks dumped trash, it clogged conveyors. A cooling fan didn't work. A trash feeder system jammed. When one part of the plant didn't work, the whole thing would have to shut down.

When garbage moved to the second stop in the plant, a futuristic tepee-shaped storage bin, it turned almost as solid as rock when stacked up for any period of time.

But the plant always worked just well enough to appease city, federal and company officials. Kuchta said that "I'd meet with private industry people and they'd say, 'You've got to have faith. You're bound to have some problems, but they can be ironed out."

There were doubters from the start. Monsanto was newcomer to the solid waste field. According to several city and federal officials interviewed, the firm was so confident that its pyrolysis process would work that it ignored developed technologies.

"Monsanto was confident it could deliver. They didn't lie, or play any hanky-panky games in the early days," one EPA official associated with the project said. "When problems developed, they made more promises. After a number of these promises you get annoyed. They reached a point in time when they ran out of expertise."

Almost everyone involved in the plant agrees on what went wrong: scaling up the plant from 35-tons-a-day capacity to 1,000 tons a day was simply harder than anyone thought.

"We thought scale-up chances were very good," said D.L. Chapman, a Monsanto vice president. "We're as disappointed as the city that the project didn't turn out as well as we'd hoped."

After months and months of making costly modifications to the plant, the bottom began to fall out early this winter. Begenning in November, Monsanto made three attempts at conducting 30-day test. Each failed.

Then, in December, a private consultant, hired by the city, concluded the plant couldn't ever operate properly in its present form. The consultant, William M. Harrington Jr., said the city could either spend $12 million to modify the plant, or scrap it, salvage $6 to $8 million in scrap rori, and build a traditional incinerator. This would cost about $11.3 million, he said.

The final blow came on Feb. 3 when Monstanto vice president Chapman recommended the project to be terminated.

The city refused, and the private firm pulled out a week later. The city has operated the plant since, and it has worked better and longer than ever before. As of last week, it was handling an average of 530 tons of trash daily, and selling an average of $5,200 worth of steam to the Baltimore Gas & Electric Co. daily. This is still well below projections.

"The plant is working. It's producing steam," press spokesman James Kaplan declared Friday. "But we still have severe problems we have to resolve."