SNOWPURE, LLC, 2005-2018
VERSION 3.5 (XL+EXL)
–FEBRUARY 2018
PAGE 26
3. Maximize the removal of silica and boron by the RO
If the total ion load to the EDI is lowered, the working section of the module will be small,
and the polishing section relatively larger. The larger polishing section will aid the
removal of the hard-to-remove ions.
CO
2
can be removed by the RO if the pH of the feed is raised (the pK
1
of carbonic acid
(H
2
CO
3
) is 6.35). Hence, with moderately high pH, bicarbonate ion can be removed. Of
course, hard cations (Ca
+2
, Mg
+2
) must be removed first to operate the RO at high pH.
CO
2
can be removed as a gas after the RO (see Liqui-Cel
®
), which is helped with low pH
(so the CO
2
is not in ionic form).
The pK
1
of silicic acid (H
2
SiO
3
) is 9.8. The pK
1
of orthoboric acid (H
3
BO
3
) is 9.3. Only
with pH > 10 are silica and boron charged. This is the theory behind the HERO
tm
membrane process.
In an EDI system, raising the pH of the feed is counter-productive. The addition of
NaOH before the EDI simply raises the ion load for the working section of the EDI, and
the pH returns to 7.0 by the end of the working section. The size of the polishing section
is now smaller. Therefore, we do not recommend raising the pH of the EDI feed.
See Appendix #8 for a more in-depth discussion of silica removal.
Temperature
Pressure Drop vs Temperature
Pressure drop depends on temperature mostly due to the effect on the viscosity of
water. The table below shows the absolute viscosity of water (cP) at temperatures of
interest, and the relative viscosity (based at 25°C). Pressure drop will increase or
decrease proportionally to the viscosity(note that at 5°C the viscosity of water is 70%
higher than at 25°C).
Temperature
Relative Water
Viscosity
Pressure
Drop Factor
5°C (41°F)
+70%
1.70
15°C (59°F)
+28%
1.28
20°C (68°F)
+12%
1.12
25°C (77°F)
0%
1.00
30°C (86°F)
-10%
0.90
35°C (95°F)
-19%
0.81
Stack Resistance vs Temperature
As temperature increases, the resistance of the stack will decrease. At a given voltage,
current will increase. One cause of this phenomenon is increased ionic activity at higher
temperatures. All other things being equal, the stack resistance will change about 2%
per 1
C.
The quality optimization will depend on other factors (below), and so the optimum setting
of the voltage will change with temperature.