Pilot testing concept
The authors" company—Energy Recovery Inc.—has demonstrated osmotic power generation using a reconfigured saltwater reverse osmosis (SWRO) process. The scheme was that shown in Figure 2, but with no freshwater circulation or turbine. Circulating fresh water was not possible because the tight permeate carrier-envelope construction of currently available SWRO membrane elements does not allow crossflow on the fresh water side. In lieu of a turbine, a portion of the dilute saltwater from the membranes was discharged through a throttle valve.
The two water sources used for the test were reverse osmosis permeate and Pacific Ocean saltwater. The saltwater had a total dissolved solids concentration of about 34,000 milligrams per liter ion at 57F and an estimated normalized osmotic pressure of 393 psi. Also used were 21 eight-inch-diameter spiral SWRO membranes (FilmTec SW30XLE-400i) from Dow FilmTec Corp. (www.dow.com) with a total nominal surface area of about 8,400 ft2. The SWRO process had a capacity of 66,000 gallons per day. Commercial pumps were used for saltwater supply, saltwater boost, and fresh water supply. Energy Recovery's PX-70S Pressure Exchanger device served as the ERD shown in Figure 3.
Figure 4 illustrates the system performance during one of the demonstration test runs, identified as "Test 4" in Table 1. When saltwater was exposed to fresh water across the membrane, the saltwater pressure rose to approximately 320 psi, or about 81% of the osmotic pressure. This pressure fell to about 260 psi (81% of the initial static pressure or 66% of the osmotic pressure) when the discharge valve was opened. The valve was adjusted to maintain a steady permeate flow rate of about 9 gpm, corresponding to a normalized membrane flux of 1.7 gallons per square foot of membrane area per day (gal/ft2/day). Saltwater was delivered by the booster pump at 58 gpm. The ratio of permeate flow to membrane feed flow (the dilution rate) was 13%. The saltwater pressure fell with time until the test was stopped after 10 minutes.
5. More pilot tests. Osmotic power demonstration Test 5 results. Source: Energy Recovery Inc.
Figure 6 compares the membrane performance during both test runs. In Test 4, the membrane power density began at about 1.8 x 10-4 horsepower per square foot (hp/ft2), or 1.5 watts per square meter (W/m2), and decayed by half in about 5 minutes. In Test 5, the membrane-specific energy began at 0.8 x 10-4 hp/ft2, or 0.65 W/m2, and decayed by half in less than 15 minutes.
6. Pushing the membrane. Membrane-specific energy performance is a key design parameter. Source: Energy Recovery Inc.
Initial system performance data are presented in Tables 1 and 2. In both tests, the system generated hydraulic power, but at a net loss after accounting for the pumps" energy consumption.
Table 2. System performance data from Test 5. Source: Energy Recovery Inc.
First results
This demonstration of osmotic power production successfully exploited the osmotic pressure between saltwater and fresh water to produce a pressurized stream of diluted saltwater that could be used to drive a turbine. However, the tests consumed more power than they created, indicating losses or inefficiencies in the process. Specifically, three types of losses are evident:
- The difference between the initial pressure and the osmotic pressure of the saltwater.
- The pressure drop through the membrane.
- The decay in operating pressure over time.
When the saltwater was initially exposed to fresh water across the membrane, the saltwater pressure increased as osmotic flow occurred. However, before the saltwater booster pump was started, there was no circulation or agitation of the saltwater, so dilute saltwater near the membrane reduced the osmotic pressure across it. This reduction is seen as the difference between the osmotic pressure of the saltwater (about 415 psi) and the initial static pressure of the saltwater (320 psi in both tests). Once the booster pump was started and the discharge valve was opened, the crossflow of saltwater across the high-pressure side of the membrane elements reduced the accumulation of dilute saltwater in the vicinity of the membrane surface.
The sharp reduction in saltwater pressure after the discharge valve was opened is evidence of frictional losses associated with permeate flow. The difference in permeate rates was insufficient to be detected as differences in frictional losses between the two demonstration tests. Higher permeate frictional losses would be expected at higher flux rates.
The decay in operating pressure with time (Figures 3 and 4) indicates that osmotic pressure fell as salts diffused through the membrane to the fresh water side. Any increase in salinity near the membrane surface would accumulate and reduce the osmotic pressure due to lack of circulation or agitation of the fresh water. This effect would be minimized by a properly designed osmotic power membrane, which would allow crossflow on both sides.
Email
Comments
Print
Reuse
