Posted by nicksanspam on July 5, 2004, 10:10 am
I wouldn't use the word bogus. Engineers are so disagreeable... :-)
I'm thinking that's the mode, say 90% of the time...
It might help to consider the "counterflow" aspect more.
Say 2 sections of tubing and 2 volumes of greywater...
If each section of tubing holds 6 gallons, RC = 6x8.33/(78.5/2x10)
= 0.13 hours. Suppose we use water in 6 gallon bursts every 2 hours,
and call the temps Tfl (lower fresh water), Tfu (upper fresh water),
Tgl (lower greywater), and Tgu (upper greywater)...
20 TFLP:TFUP'initialize upper and lower fresh water temps
30 TGL0:TGU0'initialize upper and lower greywater temps
40 FOR D=1 TO 5'days
50 FOR H=0 TO 22 STEP 2'hours
70 TGU=(15*TGU+6*100)/21'initial upper greywater temp
80 TGL=(15*TGL+6*TGUOLD)/21'initial lower greywater temp
90 TFU=TFL'move lower to upper fresh water, then wait 2 hours
100 TFL=(6*50+21*TGL)/27'lower fresh and greywater temp after 2 hours
120 TFU=(6*TFU+21*TGU)/27'upper fresh and greywater temp after 2 hours
140 IF (D=1 OR D=5) AND H<6 THEN PRINT D;H;TFL,TFU,TGL,TGU
150 NEXT H
160 NEXT D
1 0 88.88889 88.88889 88.88889 88.88889
1 2 80.24691 91.35802 80.24691 91.35802
1 4 75.99451 90.80933 75.99451 90.80933
5 0 66.66674 83.33341 66.66674 83.33341
5 2 66.66672 83.33339 66.66672 83.33339
5 4 66.66671 83.33337 66.66671 83.33337
Looks like e = (83.3-50)/(100-50), ie 67%. Not too promising.
Would it get better with more sections or smaller time steps?
(66.66671+83.33337)/2 = 75.00004. How bogus :-)
Thanks for the faint praise. Got any constructive suggestions?
Posted by daestrom on July 5, 2004, 2:24 pm
More sections. The more you keep the greywater segregated into separate
regions, the better. Keep the 'upper' and 'lower' greywater separate with an
insulating barrier except for a small flow path to let the water through
when running. The more sections you break it down into, the better. In the
extreme, you would have two concentric tubes, one for grey water flowing in
one direction, and the freshwater flowing in the other tube in the other
direction. Do this with 300 ft of two different tubes (say 1" inner and 1
1/2 inch outer tubing. If you coil it up to fit in a convenient space,
insulate between successive turns. This would be the 'classic' counter flow
heat exchanger. You want the two volumes to be as close as possible to each
other, so depending on the wall thickness you may need to adjust the size of
the outer tubing. If the wall thickness is 1/8 inch with ID of 1" and OD of
1.25", then you might use 1 7/8 for the outer tube (if you can find that
No one doubts that a higher effectiveness can be achieved, it just gets more
and more bulky/expensive. But the tube within a tube idea, it should still
be pretty cheap (add cost of second tube, but remove cost of 55 gal drum).
Coil it in any frame. Insulate each layer from the next with a circular
piece of foam with a notch cut in the edge for the tubing to come up through
it in one spot. And of course insulate the outside much like you've already
described. Ideally, the inner tube would be centered within the outer, but
*that* could get tricky. If it touches the outer on one side, it should
still be pretty good.
With 300 ft coiled in 2 ft diameter circles of 1 1/2 inch tubing and 1 inch
foam between them, it would be about 48 coils stacking about 10 ft long. Th
e water weight would be less than 228 lb, so even with the tubing and
fitting weight it should't be *too* heavy to put just about anywhere as long
as it's laying on its side. Standing up, the weight might be a bit too
concentrated for a frame-wood floor for extended time (although I 'weight
test' my floor to 270 lb every time I stand up. But then, I don't stand in
one spot for months ;-)
Maybe add a wye-strainer to the grey water inlet to keep some things from
fouling the outer tubing (depends on exact nature of 'grey'). And/or
disassemble for cleaning once in a while when the temperatures start to
degrade (monitor inlets/outlets once or twice a year to get a feel for
this). Or just a bottle of drain cleaner once in a while.
Ya, well perhaps the single precision floating point math of your BASIC
program has some round off errors in it. Bet the *real* answer is
(66.6666..... + 83.3333....)/2 = (150)/2 = 75.000000....... ;-)
Engineers also know the limitations of their tools :-)
Posted by nicksanspam on July 5, 2004, 5:36 pm
More time steps might help. It takes 6/1.25 = 4.8 minutes to fill
half the tubing, which is comparable to its time constant.
I hope that isn't required, because it might add to clogging problems.
Greywater will enter the top of the drum via a horizontal part of a
1.5" elbow and leave from the bottom via a 1.5" elbow on a dip tube.
I'm hoping it will stratify, between bouyancy forces and poor downward
heatflow. With no bouyancy and 0.00554 ft^2/h thermal diffusivity, 60 F
water with a 100 F surface might warm to 80 F 1' below in 180x1^2 hours.
I thought about that, but the outer tubing becomes more expensive, and
there's the clogging problem.
This kind of tubing will bend to a 20D radius without kinking...
Hmmm. Hair, soap, food particles. Some septic tanks use nylon stocking
output filters. Input filters... yuck. Tiny crud particles might still
agglomerate inside the heat exchanger.
Posted by daestrom on July 6, 2004, 9:10 pm
Would give you a smoother curve, but would not change the end point.
Yes. The ideal case is each layer does not mix with another, nor is any
heat conducted between sucessive 'grey' 'layers'. Since water doesn't
really behave perfectly in a large tank, we break it up into several small
ones with insulation between them.
Save cost of drum, that helps alleviate the extra cost. Yes, there is the
clogging problem. But even with the large drum, there would be a
sedimentation problem. The bottom would fill with 'silt' and such. Just
could operate longer before needing to clean it out.
Well, a wye strainer would be a lot easier to flush out once a month or
whatever. And so the amount of 'crud' that enters into the heat-exchanger
would be less and *it* would last longer before needing a thorough cleaning.
Either a high-pressure/high-flow flush, or disassembly. If expect to
disassemble often, could take that into account with the fittings/design.
Empty, disconnect, take outside, pull inner tube from outer,
scrub/brush/rinse, reassemble, reinstall. If it's only once a year or less,
might be a not-bad summer weekend project.
Of course, the commercial version of GFX heat exchanger doesn't require this
maintenance, but is more expensive ;-)
Posted by nicksanspam on July 7, 2004, 12:06 am
I disagree. Seems to me more time steps would show more stratification.
Can't get much better than 1' in 180 hours, ignoring bouyancy...
How will the water behave without insulation?
I've collected about 300 free drums. Floorspace is also an issue.
...ie a 2x20x1.5 = 60" diameter.