Volume 42 Number 7
RO: New Device Shatters Seawater Conversion Conceptual Barriers
Photographs, figures and/or graphics that may illustrate this article are visible in the printed version of the article only. To receive a copy, please make a request at email@example.com. Be sure to include the article title, author(s) name(s), the issue, your name and your fax number or full address in the email.
What happens if high recovery design is no longer crucial -- Is this an energy recovery device or a pump?
"If we could ever competitively, at a cheap rate, get freshwater from saltwater, that would be in the long-range interest of humanity (and) would dwarf any other scientific accomplishments"
President John F. Kennedy
April 12, 1961
Summary: A large reason why conversion of seawater to potable water has been impractical on a global scale has been the energy cost required to operate the plants. But what if that cost could be drastically reduced by recapturing energy inherent in the system? That possibility is explored here.
Desalination, or production of potable water from seawater, has been done for centuries. Until now, however, desalination had been considered a costly alternative because of the high energy consumption in the process.
Desalination and membranes
Early desalination plants relied on evaporation technology. The most advanced seawater evaporation desalters using multiple stages have an energy consumption of over 9.0 kilowatt hours per cubic meter (kWh/m3) of potable water produced. For this reason large seawater desalters were initially constructed in places with cheap, abundant energy such as the Middle East or next to process plants with available waste heat.
In the 1970s, a new process called seawater reverse osmosis (SWRO) was developed which made potable water from seawater by forcing it under high pressure through a semi-permeable membrane thus separating out salts and impurities. These salts and impurities come out of the SWRO device as a concentrated brine discharge stream.
The brine stream of a SWRO contains a large amount of pressure energy that can be recovered with the right device. Absent a good energy recovery device -- and because of membrane quality and pressure drop limitations -- many SWRO plants built in the '70s and early '80s had an energy consumption of over 6.0 kWh/m3 of potable water produced.
Search for energy savings
In 1985, Dow Chemical division FilmTec developed the first commercial low-pressure, single-stage SWRO element. At the same time, pump manufacturers were adapting existing technology such as reverse running turbines and Pelton wheel devices to SWRO plants. A Pelton wheel is a type of water wheel invented in San Francisco during the 1850s Gold Rush to use water pressure from Sierra Nevada streams as an energy source. The device consisted of a nozzle to convert water pressure into kinetic energy contained in a high velocity jet. This was directed to a series of buckets -- or metal vanes in the modern sense -- around a rotating shaft that intercepted the jet stream and converted the kinetic energy into rotational energy to turn a shaft.
New membrane technologies and first generation energy recovery devices based on this principle made possible seawater desalination with an energy consumption of just below 4.0 KWh/m3. But rotating machinery of first generation energy recovery devices adapted from other uses to seawater conversion were made of metal parts that often exhibited high corrosion and wear maintenance problems when placed in a marine environment.
By 1990, a second generation of energy recovery devices came to market that used corrosion resistant alloy parts such as 904L stainless steel. At about this time, the hydraulic turbo charger was also developed. This proved more reliable and had lower maintenance but still suffered from the limitation of only being able to recover 50-to-80 percent of the energy in the high-pressure brine stream from SWRO plants due to various inefficiencies (see Figure 1).
Efficiency double dip
The typical Pelton wheel arrangement used to recover energy in a seawater reverse osmosis plant has a fairly high energy efficiency -- on the order of about 80-to-85 percent. However, one characteristic of Pelton wheel, reverse-running turbines and other energy recovery turbines (ERTs) is, in SWRO applications, they usually transfer the recovered energy from the brine through a shaft or belts back to the primary high-pressure pump.
For this reason, the loss in energy efficiency of Pelton wheel and other ERT devices must be added to the energy losses of the pump it's driving. This "double-dipping" of efficiency penalty means, as the permeate recovery of an RO unit decreases, more high-pressure brine is sent to an ERT device, resulting in a rapid decrease in the overall energy recovery efficiency.
Figure 2 shows the energy efficiency of a typical RO plant with a centrifugal high-pressure pump equipped with Pelton wheel ERT operating at 950 pounds per square inch (psi) as the water recovery percent changes. The figure illustrates why most SWRO plants built over the past 20 years have been designed to operate at very high recovery. Simply stated, a SWRO plant equipped with conventional energy recovery devices is about twice as efficient when working at 45 percent recovery than when working at 25 percent recovery. This drove RO plant designers towards seeking higher and higher recoveries.
Flow imbalance issue
Product water flow, or permeate, rate is one of the key RO design parameters. The type and source of the seawater dictates the membrane pretreatment required to minimize membrane fouling.
Osmotic pressure refers to the pressure and potential energy difference existing between two solutions on either side of a semi-permeable membrane because of the tendency of water to flow in osmosis. Every 100 parts per million (ppm), or milligrams per liter (mg/L), of total dissolved solids (TDS) generates about one psi of osmotic pressure. This osmotic pressure must first be overcome by water pressure for an RO membrane to become effective. In the case of seawater, which contains from 30,000-to-50,000 ppm TDS depending on the sea, the osmotic pressure can be in the range of 400 psi.
An RO plant is a seawater concentrator. When it's operated at very high recovery, the salinity of the seawater being treated increases rapidly from the first element in a membrane array tube to the last element of the membrane array tube. This means the last elements in the array are exposed to much higher osmotic pressure and thus have much lower net driving energy to do their desalination work.
Let us examine what happens to individual RO elements in a tube containing 6 RO membrane elements in series when an SWRO plant is operated at 45 percent recovery. This is illustrated in Figure 3.
What happens as a result is that the first element in the array is producing a high permeate flow rate and the last element is hardly working. This imbalance means the first elements are operating in a range where they tend to foul rapidly, while the last elements are producing very little water for the investment.
Also at this high recovery, downstream RO elements produce permeate which may not meet World Health Organization drinking water guidelines (>425 ppm TDS). For these reasons, SWRO plants -- to produce potable water from high salinity seawater (such as Arabian Sea water) in a single pass -- have to work at a higher pressure or lower recovery.
One excellent line of SWRO plants built in Spain for Arabian Sea -- Persian Gulf -- applications is designed to work at 25 percent recovery (see Figure 4). Keep in mind the salinity in the Persian Gulf is typically 49,000 ppm, compared to the Caribbean Sea which is about 32,000 ppm.
However, as we saw in the energy graph (see Figure 3), the energy consumption at this recovery when using conventional technology is over 6 kWh/m3.
Over the last two decades, various inventors have tried to develop advanced energy recovery devices to overcome the inefficiencies inherent in SWRO because of low recovery, "double-dipping" and high-pressure operation. Such devices are based on the principle of either double-sided pistons or the isobaric chamber, which permits transfer of pressure energy from one stream to another in a piston-less closed chamber.
The general idea of these devices is to transfer the pressure energy from the outgoing brine into the incoming feed directly -- without using turbines or pumps. The challenges are many, including valve and timer designs and minimizing mixing of the fluids. Mixing of brine with the feed is undesirable because it increases the feed salinity.
Early devices used a combination of pistons or bladders as well as valves and timers to try to recover this energy. Some devices worked well initially, but suffered high maintenance problems. Others were fitted with artificial intelligence programs only to suffer early demise in an industry where the prevalence of unskilled operators usually calls for simplicity.
(See Figure 5.)
In 1992,another company was launched around the simple idea that a spinning ducted rotor made from engineering ceramics could transfer the pressure energy from the SWRO brine to the incoming feed stream. The idea evolved into a 4-inch diameter x 4-inch long commercial pressure exchange device with no bladder or piston separating the two streams. Rather, the rotor is spinning at 1,500 revolutions per minute (rpm); therefore, residence time is so short very little mixing is possible (<2 percent). The amount of time for the water to pass through is roughly 1/30th of a second -- which, for comparison, is the slowest shutter speed possible on a 35-mm camera to shoot a handheld image.
Figure 6 shows the ceramic rotor of this commercial device. Since it actually transfers energy directly from the brine to the seawater feed without any "double-dipping" caused by multiple rotating shaft inefficiencies, the actual efficiency is roughly 95 percent within a broad flow range. This is best illustrated by repeating our efficiency curve for an exchange unit recovering the energy from an SWRO plant fed with a centrifugal pump as compared to a Pelton wheel driving an identical pump under the same conditions (see Figure 7).
As you can see, the efficiency curve of the exchange unit is very flat. This means the RO plant designer no longer has to worry about recovery efficiency driving his design. Now, product quality, flow balance and pretreatment equipment size become the overwhelming design criteria.
Reaching theoretical limits
So, is seawater desalination for 2.0 kWh/m3 (7.6 kWh/1,000 gallons) still impossible?
The early commercial SWRO applications were mostly in energy rich desert nations or in Islands. In most islands, energy costs tend to be very high because of relatively smaller scale distribution and storage. In St. Maarten, for instance, electricity is $0.145/kWh. If the SWRO energy consumption is over 4 kWh/m3 using traditional technology, power is 40 percent of the water sell price and becomes an overwhelming part of the desalination cost structure.
Right now, as we speak, there are over 50 pressure exchange devices as described in this article installed in commercial plants. Using this technology It's possible to design and operate SWRO plants with Caribbean seawater at 780 psi and 36 percent recovery -- and have a process energy consumption of less than 2.0 kWh/m3.
Energy recovery or pump?
Because the exchange unit transfers high-pressure energy directly to incoming seawater, the device also acts as a high-pressure pump. The high-pressure pump may now be sized to just match the permeate flow. At the 45-to-50 percent recovery range of most existing modern SWRO plants, incorporating pressure exchange units in their design allows engineers of large SWRO plants to eliminate half the high-pressure pumps from their plants.
Before pressure exchangers, for instance, the designer of a 12,000 cubic meter per day (m3/day) plant might incorporate six 2,000 (m3/day) control blocks each with an 800 gallon per minute (gpm) high-pressure pump. The designer now has two choices. If he designs with 2,000-m3 trains he now needs only six 375-gpm high-pressure pumps. Or he can use the same 800-gpm high-pressure pumps and cut the number of control blocks down to three. This halving of the main high-pressure pumps means the exchange unit design has a significant capital cost advantage over conventional designs.
(See Figure 8.)
Advances in reverse osmosis membrane technology and reduced energy and capital costs mean that -- for the first time ever -- it's possible to produce potable water from seawater at a cost of less than $1 per cubic meter in many locations worldwide. Although the cost may be a bit higher on islands or in areas with high power costs, the pressure exchange device described here has the potential to rapidly expand the market for seawater desalination.
About the author