There's always been a hole in the plan to turn the rock-mining pits of West Miami-Dade's "lake belt" into gargantuan reservoirs for Everglades restoration.

There are a slew of holes, in fact, along with countless cracks and fissures in the walls of the limestone quarries, which make them about as watertight as kitchen colanders. Not a good quality for storing water.

After much study, federal engineers have narrowed the search for ways to reduce the leakage. The potential fixes display real out-of-the-box thinking. Out of the icebox, in one case.

Something called "ground freezing" made the official first cut with four other technologies under consideration by the U.S. Army Corps of Engineers. That would entail bringing in industrial refrigeration units and sinking probes as deep as 120 feet underground to create a frozen wall of rock, soil and water as thick as 30 feet.

The idea of putting subtropical Florida on ice has environmentalists shaking their heads. They're equally dubious about other options: plastic lining, spray grout, cement-like "slurry" or the more conventional notion of pilings.

"They're all really as insane as freezing water in Florida," said Barbara Lange, Everglades cochair of The Sierra Club. A year ago, The Sierra Club and other environmental groups sued the Corps to halt further expansion of rock-mining into a 21,000-acre region designated by the state as "the lake belt."

The Corps defends its approach, saying it simply exercised engineering diligence in exploring the widest array of possible seepage stoppers. It expects to make a decision within months on what to test for a $23 million pilot project.

If an underground ice wall isn't a practical solution -- and the Corps agrees it's probably not -- one of the other technologies may be, said John Keiser, the Corps' project manager for the lake belt pilot project.

"I guess anybody can make a joke out of anything," Keiser said. "Maybe ground-freezing is at the far end of the scale, but we'd rather at the end of the day know we went through a thorough analysis."

Tom MacVicar, a consulting engineer who represents the mining industry, said the Corps' critics were unfairly making sport of an assessment process they don't understand.

"They throw everything that can possibly work or can be used on that list," he said.

BOLD IDEAS

The key goal of the $8 billion Everglades project is to restore the natural flow of water through the River of Grass. One of the biggest challenges is ensuring that there's enough water to do that job while still supplying the growing needs of farms and cities.

To do that, the Corps, which is managing the restoration, has devised some controversial and expensive new water storage ideas -- some 330 underground wells and three "in-ground" reservoirs, or giant rock pits. The largest two reservoirs, 4,500 and 5,200 acres, are planned just west of the junction of U.S. 27 and the Florida turnpike extension.

Environmentalists fear both plans could taint the supply of pure underground drinking water. They say the reservoirs carry a staggering price tag, $1.2 billion, pose a threat to Miami-Dade County's drinking water in nearby well fields, and could suck precious water from the Everglades just a few miles west, harming efforts to restore natural flows and levels.

Compounding problems is the fact that the Corps isn't developing an alternative if the pits prove too porous, said John Adornato, regional representative for the National Parks Conservation Association, another group in the lake belt suit.

"They're putting all their eggs into one basket, and they don't even know if the basket has a bottom yet," he said.

But the Corps said that's exactly what the test project is designed to do: evaluate whether the pits can hold water well enough before investing in larger-scale storage.

The reservoirs have some advantages for water managers.

They will hold huge volumes of water that could be used during droughts and would utilize pits created by commercial mining operations.

QUESTIONS

But the Corps admits there also are questions about costs, water quality impacts and whether it's even possible to stop the natural underground water flow through South Florida's porous bedrock.

Keiser, the project manager, said the Corps weighed a wide range of possibilities and narrowed it down to technologies with good track records that could also withstand the impact of blasting used in rock-mining operations.

Ground-freezing, while never used in South Florida before, has been used to seal off underground flows elsewhere, said Keiser, and has the advantage of being easily reversible, another criteria for the test effort.

"Basically, for ground-freezing, it would be pull the plug," he said.

Still, he said, the Corps intends to weed the deep freeze from the list. High operating costs are one concern, but the more serious one is the speed at which water moves underground -- too fast to put solidly on ice.

THE CANDIDATES

Other possibilities and problems:

• Drive pilings, such as steel panels, around the pits. It's a common practice but challenging to sink so deep.

• Fill a deep trench around the pit with a ''slurry'' that hardens like cement. That option is difficult to reverse.

• Encircle the pit with a ''geomembrane,'' or plastic sheeting, then fill the trench with soft slurry. The plastic sheeting complicates installation.

• Injecting grout into rock under high pressure. That process might leave wide gaps in the seal.

Keiser said it's likely the Corps will select one or more systems to test in the next few months.

Then it has to find a spot for a pit and dig it.

"I think in some form, the technologies will all work," Keiser said.

"It's just the degree it will work."

What works best probably won't be known for years. The Corps still has to select a site for the test pit, dig it, then install and test the selected leak-pluggers.

All that is set for completion around 2009, with a final recommendation on the full project due in 2013.

The current restoration schedule doesn't call for construction to begin on the larger reservoirs until 2018.