Membranes

Jerry Knippel doesn’t mind helping you learn from his mistakes. As division director of water for Dunedin, Fla., he is really kind of used to telling people about the city’s experience with membrane technology, as a speaker at conferences or in conversations with his peers.

It is not that this experience has been bad. Indeed, the city considers its 9.5 mgd reverse osmosis (RO) water treatment facility, brought online in 1992, a successful project.

It is just that the use of RO and other membrane processes, though expanding, is still fairly uncommon in much of the U.S., so Dunedin’s experience draws a lot of interest.

Both water quality and quantity concerns in the late ’80s led Dunedin to look at expanding its treatment capacity, with conventional lime softening and membrane softening as its two options. After determining that the costs for each option were comparable, the city chose the latter for two fundamental reasons: A plant using membrane processes would allow Dunedin to expand its use of brackish water as raw water; and the high-quality finished water produced by such a plant would have the greatest chance of meeting any future, more stringent federal or state regulations.

RO is one of five basic treatment processes using membranes, along with nanofiltration (NF), ultrafiltration (UF), microfiltration (MF) and electrodialysis reversal (EDR). Simply put, the first four of these involve forcing water through hollow-fiber or spiral-wound membranes to remove various types of contaminants. In an EDR process – rarely used in municipal plants – a direct electrical current is passed through membranes to draw out dissolved ions like sulfate.

MF and NF operate at low water pressure and remove relatively large-sized particulates in raw water, including Cryptosporidium and Giardia. RO and UF involve tighter membranes and much higher pressure, and thus are able to remove particulates as well as dissolved solids such as calcium carbonate.

“Looking at it very simplistically, sometimes we joke and say that membranes are dumb; they only know how to do one thing: let clean water through and reject the rest,” Knippel says.

For Dunedin, like several other membrane users on Florida’s coasts, the desire to use more brackish water meant it needed RO’s ability to remove just about everything. Tapping this brackish water would allow the city to continue using only its local wellfield, rather than buying water from surrounding counties or the regional authority as demand increases.

“We [want to] remain self-sufficient, and that’s why our local water resources are so important to us,” Knippel says.

The city also wants to remain in compliance with water standards. Like most other public water suppliers, Dunedin has been concerned with rules regarding contaminants like synthetic organic compounds, microorganisms and disinfection by-products (DBPs), and decided RO would be the most effective in meeting regulations that promise to get tougher (at many “traditional” facilities, suppliers are faced with limiting DBPs while still using enough chlorine for disinfection and maintaining a residual in the distribution system).

Knippel says some utilities may have a problem at the end of 1998 when federal rules are set to lower the maximum amount of DBPs allowable from 100 ppb to 80 ppb.

“I can tell you right now, there are a lot of utilities that are barely making it at the 100-part level,” he says.

By using membranes, on the other hand, suppliers can remove naturally occurring organics in raw water before they ever reach the disinfection stage to be mixed with chlorine, thus eliminating half of the DBP equation.

“The beauty of the membrane technology is that it removes the organic precursors before they have a chance to meet with the chlorine,” Knippel says.

More and more suppliers are finding this kind of capability appealing, and serious interest in membranes has grown tremendously in the past few years, according to Steven Duranceau, senior engineer with Orlando, Fla.-based Boyle Engineering. In addition to concern over existing and impending regulations, a steady decline in costs and continued advances in the technology – like lower operating pressures – are also driving Membranes also have in their favor a small footprint and a modular nature, so that new membrane skids can be added to meet increased demands for water. And, when they must be replaced, new membranes can be installed to take advantage of advances in technology, although Duranceau cautions that idea “really hasn’t played itself out, because most plants are just now entering the period of membrane replacement.”

Still, the reasons for at least exploring membranes have begun to outweigh the reasons not to in a lot of situations.

“[Use of membranes] used to be pretty much limited to Florida and California, and now you’re finding it in virtually every state,” Knippel says. “It’s really finding its way throughout the U.S.”

Duranceau, who has chaired AWWA’s membrane processes committee for several years, says this year’s edition of the membrane conference, the fourth since 1991, drew more than 600 people from 27 countries.

And it is not just larger systems showing such interest. As costs have come down, smaller systems are considering the technology as well.

“Smaller systems are beginning to realize that with membranes, they can treat a wide variety of things,” says Dallas Post, AWWA’s operations division and small systems program manager.

Post says AWWA, which as recently as three years ago did not actively promote the technology, is now a clear proponent. He says that numerous pilot studies have been underway at utilities, and the results coming in are worth a close look.

“I think that [membranes] really deserve serious consideration by any utility, regardless of their water source,” he says.

Likewise, Knippel says that “I really can’t think of a situation where membranes could not do a very effective andefficient job of treating a water supply.”

That is not to say building and operating Dunedin’s RO plant has been problem-free. After getting past a number of obstacles in embracing a technology not widely used, he offers the following points of advice:

* Since utilities each use raw water of different quality and thus may have different treatment goals, pilot testing is critical. Testing should be conducted early in the process and may have to be run throughout the year in case problems are seasonal;

* Beginning the permitting process for concentrate disposal should be one of the first steps. Concentrate – water that carries the contaminants removed in the membrane process – is regulated at all levels. Dunedin, for example, currently pipes its concentrate to the city’s wastewater treatment plant;

* Utilities leasing a pilot test unit should be sure the manufacturer does not include language in the lease that names itself the exclusive supplier for the actual project;

* The supplier or manufacturer of a new facility’s membrane equipment should be required to install the process instrumentation and controls, for the sake of preserving membrane warranties;

* Utilities should hire a design consultant with a proven track record in working with the technology. “As the technology becomes more popular, there are more and more engineers and consultants who are qualified to effectively design such a facility,” Knippel says. “Initially, that wasn’t necessarily the case”;

* The contractor should be required to operate the facility during start-up and the first 30 days, and to be present for the first week that city personnel take over. Personnel should be hired as early as possible; and

* Even though the plant is new, a preventive maintenance program should be established at the start.

The list of lessons other suppliers could add to the list will grow if the predicted boom in membrane use takes place. But if the technology fulfills its promise, consumers of public water supplies stand to be the winners.