Pump mystery

pex?
In looking more deeply at your problem, two possibilities emerge.
background/line of thought, two paths:
Generally, if a hose is rated for intake (inside less than atmospheric pressure), it has two or more layers. If they separate, they block flow: e.g., either a separation of the linings from each other so the inner lining loses some of the outer-lining bond-support and is weaker such that the inner can collapse in a range of less-than atmospheric pressures, or the inner lining tears across the longitudinal axis of the hose and it acts like a checkvalve.
For example - Hose A is longer than hose B, and both hoses are connected to like intake tanks, i.e., effectively a common intake - let hose B, the shorter hose, have an internal separation
When the pump draws from both hoses, note that the pressure along the hose is a function of the pump and the flow - having twice the flow when two are connected means half the effective pressure in each (ref power equation for flow and pressure with fixed power pump) , but still more than enough pressure to cause the separated hose inner wall to collapse/fold-under-flow in the hose and block flow. (i.e., because the pump only has so much power available to create the pressure in this case, think a pump intake on one small line vs the pump intake on a hundred parallel small lines. Or just look at the horsepower equation for flow and pressure and think fixed horsepower. If flow goes up, pressure has to drop or the pump has to get bigger)
Then when the pump draws from one hose, the pressure increases and the fold then turns on itself, folding back out of the way.
Only problem with this scenario, and any scenario that drains one tank before another when they are supposedly commonly connected, is that the first-drained-tank's hose should fill with air when its tank is empty. That means that the commonly connected pump has to 1) somehow create vacuum to lift the second tank's fluid when it prefers to pump make-up air still entering the now-empty first tank, or 2) not loose fluid in the first-emptied tank hose but yet block that flow, after just being the preferred flow, or 3) the tanks have pressure-relief caps that are set somewhere between the pressure created by one flow and the pressure created by two flows (or the pressure/flow needed to turn/collapse a separated inner wall. (having caps that permit no make-up air is unlikely)
Which leads to the second path -
try swapping tank caps and see if the filled-tank preference switches with it. If so, the caps are different in their pressure-intake settings and are not allowing air into the tank properly/evenly. That is, the cap with the higher intake setting is filling second (I think?)
then try it with the caps off (loose). If they still empty in sequence, that will remove the caps as a cause. And then it is likely a hose wall-to-wall separation/break.
fwiw...

Thanks for the ideas, but the hose is rigid, single layer, rigid, single layer, rigid.... Paul
hob wrote:

Dear Paul Mathews:
I feel your pain.
However, the only PVC hose I am aware of with spiral winding is *two* layer, with the spiral winding between the layers. This prevents material compatibility issues with the metal of the spiral. Especially likely to separate if heated to insert hose fittings, since the metal and the plastic have vastly different coefficients of thermal expansion.
We aren't harping on this because it is fun for us to do so. Some of us have been there.
You are going to have to get dirty, if you want this "fixed".
As others have pointed out, you can take another hose from tank-to-tank, and form a permanent siphon. That way even a small variance in tank level will start a flow to re-level the tanks. 2" should be plenty big enough.
You can also look at a better style of hose: "can be cemented over PVC pipe" looks really attractive.
David A. Smith

tried the cap swap?
collapse/fold-under-flow

The mystery is not so much as to why one seemingly equally connected tank drains first, it is more why the second tank is then emptied at all.
That is, I find it more strange that the last-to-drain tank's line doesn't fill all the lines with air once the first-drained tank is empty. An empty first tank should suck air, breaking the fluid to the second tank and filling the lines with air so the pump then would see only air - unable to establish a siphon to lift the fluid in that last-to-drain tank, it would stop any fluid being lifted from the second tank at all.
Unless there are pressurizing air-make-up feeds into the tanks that aren't noted; or the caps are of a differential pressurized type like automobile gas caps and the two caps here are of different actual values where that cap difference is greater than the pressure needed to siphon tank to tank, this starts to smack an uneasy bit of troll.....
collapse/fold-under-flow

Dear hob:
It is a puzzler. That is why I asked for pictures.
Yes. It is reasonable to expect the "most likely to drain tank" to drop to a certain level, then both tanks should drain together "inch for inch", until a stinger is exposed. But then he said "this is not a flow problem".
He said it was open to atmosphere.
It is somewhat out of character for a "troller" to get irritated at having the same questions asked over and over. They usually live for that stuff.
I am thinking that this is an indication that his "biodiesel" is actually offgassing. Although why it would preferentially affect only the closest tank... maybe this tank's contents are contaminated by biology. Or alcohol.
David A. Smith

... I visualized an interlayer of spiral fabric thread, which is quite popular for moderate pressure applications.
Brian Whatcott Altus OK

"N:dlzc D:aol T:com (dlzc)" wrote in message news:TeFQf.110$Vy.44@fed1read02... If it were a flow problem, The further tank would lower first until an equalization occured in flow pressure. (lesser fluid would create lesser pressure to help fluid flow)
It seems there is a flow problem if the first tank (the one losing fluid faster) does not empty completely first causing an air intake to occur..
Once the further tank gets more empty, the pressure it pushes the liquid out with should lessen enough allowing the closer tank to finally start to flow.
It does seem to be a flow problem. simple as that Even the way it acts shows such.
Testing phase should begin: Place a flow reducing pipe on the further tanks inlet. If it equals the draining or changes its rate.. It proves a flow problem in the closer tank and the lines for the closer tank should be checked and or replaced. :)

Dear Brian Whatcott:
Yes. The same companies that make this configuration ("nylabraid"), usually make it with a light gauge stainless steel spring embedded in the same location. Made especially for suction hoses (since it cannot collapse easily), but not at all good for "moderate pressures".
Paul said on March 9th: "the PVC hose is of the rigid-spiral reinforcement type, so there is no chance of it collapsing under atmospheric pressure"
David A. Smith

No actually you have missed the point completely. What has been pointed out is that he already has what should be a siphon hose from tank to tank. The fact that it passes thru the tee shouldn't matter. Whatever is preventing siphoning now is also the the answer to his problem.
-jim

There have been a lot of posts to this thread, I just want to check some facts.
( 1 ) the tanks empty un evenly.
( 2 ) if pumping is stopped the tanks do not level out, but stay at different levels.
IF the above are true there is not a syphon action between the tanks.
If there is no syphon action there must be either a total blockage, an air lock, or a pressure diferance.
( 3 ) both tanks have free air intake to their surface.
If the tanks don't level out ( no pumping ) with their filler caps removed then a pressure diferance is not the problem.
( 4 ) the tubes at their top points are transparent, and any air bubble would be visable...
I don't always trust visual inspection, as it's possible to mistake a compleatly full tube for a compleatly empty one. the symptoms described sound like a bubble in the tube.
( 5 ) the clear tubes are suction grade PVC reinforced with wire.
I have experiance of the inner lineing / layer of suction tubes collapsing, usualy when an unsuitable product is piped through them.
If there is no syphon action it is unlikley a colapsed tube would compleatly shut of the flow.
If there is a sypon action ( no pumping ) but it's slow then a restriction must exist.
inlet tube " sealing " to tank base, colapsed hose lineing, or a partial blockage ( odd bits of string or old bolts ) are favorites.
With free flow between the tanks and a diferance in level existing... forget the biodesil as a fuel. you have the sight of the worlds first anti gravity perpetual motion power station... :-)

Dear Jonathan Barnes:
The pump suction is at about 80% full level on the tanks. If one of the tanks is less than 80% full, siphon could break. I understand everyone's point, but the driven shaft of a pump is rarely a perfect seal.
Sounds right, be we are assured that the hoses are clear and there is no visible bubble.
If the "stingers" includes a "foot valve", to prevent having to reprime the pumps, then it is possible that the tanks will only equalize in level over some very long time.
I don't know what biodiesel looks like. Cooking oil perhaps? Not too likely it would be mistaken for anything other than what it is.
I wonder if the stinger of the "last-to-be-emptied" tank doesn't have its foot valve installed backwards (if that is possible...). Check valves have a cracking pressure...
But he describes the equivalent of a fixed differential pressure (a certain height differential) when the other tank starts emptying...
David A. Smith

Dear Jim:
No, actually I didn't: - The biodiesel is a mess to work with, - The OP is unwilling to tear the system apart.
*Add* a third hose, who's sole purpose is to siphon from the high tank to the low tank. That is *my* suggestion. Then it simply doesn't matter which stinger+foot valve (if any)+hose is actually working.
If he has foot valves in those stingers, the tanks will not equalize in any short period of time.
Remember, this pump suction is located near the top of the tanks (about 80cm on a 100 cm tank).
If one of the foot valves is bad, swapping hoses would show this. But he is so far unwilling to break the system apart to swap hoses.
David A. Smith

OK, but then you will never solve the mystery. And as far as I can tell that is the point of thread. The situation appears to present no real problem that must be solved - only a curiosity. By the way, you should be able to drain a 1000 gallon tank over night with a 1/2" siphon hose.
Yes I agree if there is a check valve in tank A that requires a certain amount of pressure difference to pop it open that would match the description of what's happening. But his complete listing of the system components doesn't include any check valves.
-jim

Dear jim:
And he doesn't say that he has installed them. I guess we'll see.
Interesting, this website (which is talking about making biodiesel, not dispensing finished product) says NOT to use PVC: We can see what other tidbits Paul will provide...
David A. Smith

Thanks for all your suggestions and ideas. The situation is not urgent at all, but the symptoms surely have me curious. By the way, there are no foot valves, no air bubbles, and the hoses are of the spiral double-extrusion type, with relatively flexible material coextruded between a spiral of more rigid material....no 2nd layer per se.
The wackiest post was the one positing 'outgassing' making a difference: the tanks are effectively open to the atmosphere. ;-)
Paul

Dear Paul Mathews:
Thanks.
Please notice that air can have the pressure dropped, and clouds will form. Note also (more to the point) that a fluid that has its pressure dropped can evolve gasses, and a pump that repressurizes that fluid will cavitate. Ever heard of cavitation? And all with "fluid reservoirs open to atmosphere."
You brought the conundurm. You don't have to act like an ass.
You can solve your own problem. Goodbye.
David A. Smith

"N:dlzc D:aol T:com (dlzc)" wrote in message news:TeFQf.110$Vy.44@fed1read02... As in debris and fungus and mold and birds open to amosphere?
Or a cap with a vent to atmosphere like on an automobile? (Which actually has a pressure differential with atmosphere)

"N:dlzc D:aol T:com (dlzc)" wrote in message news:efJQf.116$Vy.101@fed1read02... Or if there is screen on the stinger that has become partially plugged from the "open to atmosphere" entry point for leaves or molds, etc., that like biodiesel