[ExI] Deep ocean osmotic membranes

[Keith Henson]

Per recent discussion, it is expensive to clean up salt water to the
point it will not clog osmotic membranes.  So after clean up, the
standard practice is to pump the water pressure up to where about
2/3rds of it goes through the membrane leaving 1/3 of the volume as
brine.

The big advantage of going deep is that there is no reason to
concentrate the brine so fresh water is being extracted at minimum
pressure and an awful lot of slightly more salty water flows around
the membranes.  The following is a snip from the paper I wrote in 2009
on how to use energy from power satellites to make fresh water.
Graphene membranes might pass more water per unit area or be less
expensive, but  the energy needed to push the water against osmotic
pressure is the same.

It's possible that we could take advantage of biological fouling and
let the stuff that grows on the membranes *be* the osmotic membrane.

Keith

WATER

Low cost, space-based solar power will solve other problems as well.
The average flow of the Colorado River is ~630 cubic meters/sec (or
2,268,000 cubic meters/hour).  This is a useful amount.   At 1000
l/person/day, it is enough domestic water for 54 million people.  At
the minimum energy (.77kWh/cubic meter) to desalinate this flow from
ocean water, it would take 1.75 GW.  Actual energy consumption for
real reverse-osmosis desalination is about three times this amount,
largely because the discharge water has much more salt in it than sea
water.  Though SBSP produces abundant low-cost electric power, it
would be better if there were a method that would get closer to the
minimum.
One way to get fresh water from salt water is by using the ocean’s own
hydrostatic pressure.

The idea is simple:  put the osmotic membranes in deep ocean water,
the deeper the better, to recover the density difference between fresh
water and salt water.  In the deepest place in the ocean, the density
difference drives fresh water all the way to the surface.  Other
places it needs pumping assistance.  Membranes work below 300 m of
water.  Placed at 300 m the water must be pumped up the full depth.
At 3500 m, water needs to be pumped up only 200 meters.

There will be an engineering tradeoff between reducing energy cost by
putting the membranes in deeper water, the cost of pipelines to shore,
and the energy needed to drive the pumps.  An additional constraint is
the need for a cross-flow of water through the membranes to prevent
brine accumulation.

Deep bottom currents off California are in the 0.01 m/sec range.
Bottom currents in unusual locations such as the mouth of Monterey
Canyon can reach 0.5 m/sec.  High on the continental slopes 0.25 m/sec
is typical.
Not calculated is the downward flow around the membranes from the
relatively denser seawater after extraction of fresh water.  This may
be substantial.

For a design example, consider a 10 km section of freshwater collector
pipe at right angles to the current and 100 m high membranes draining
freshwater into the pipe.  This intercepts a million square meters.
At a cross-flow of 0.25 m/sec and an extraction of 1 % of the water in
seawater flowing between the membranes, the production of freshwater
would be 2500 cubic meters per second, about 4 times the flow of the
Colorado River.  Even for California, that is a lot of water, about 8
times the maximum flow of the California Aqueduct.

The power required to lift 2500 cubic meters of water 200 meters
against g would be:
P = Q*g*h/pump efficiency, where Q is in cubic meters/sec and h is the
lift in meters.  Pumps in this size are 0.9 efficient,
2500*9.8*200/0.9 is 5.4 GW.

Check, 0.77 kWh/cubic meter*2500 cubic meters/sec*3600 sec/hr = 6.9
GW, but some of the energy is coming from the density difference
between fresh and salt water.

At a penny per kWh the energy cost per cubic meter would be 5.4
GW*$10,000/GW-hr/2500 cubic meters/sec*3600sec/hr or 0.6 cents per
cubic meter, somewhat high for irrigation water, but fine for domestic
supplies.  (Capital and maintenance will add to this number, probably
not more than a cent.)

One uncertainty is biological fouling.