I think I'll put the pump between the filters. 10 micron on
the vacuum side, 1 micron on the pressure side.

Doug
s/v Callista

"Rich Hampel" wrote in message
...
After 30 years of screwing around with this stuff I cant still give a
an accurate technical reason .... my 'opinion' is the 'regime' of
particle depositionIn and the formation of 'filter cake' ..... on a
pressure filtration the deposition begins mostly on the upper surface
or at least within 5% depth of the surface, while with vacuum
filtration the deposition is essentially INSIDE the matrix of the
media. Being inside the matrix causes higher internal velocities
which drive the particles deeper and deeper into the matrix ... causing
an exponential decrease in service life. ..... its the same for depth
as well as membrane filtration.
The quandy is that the fluids are incompressible and shouldnt make any
difference due to the direction of motive pressure .... but in practice
it does, it always does.


article , Steven Shelikoff
wrote:

On Thu, 08 Jan 2004 02:51:46 GMT, Rich Hampel
wrote:

NOPE!
In pressure mode, the filter will also act as a 'coalescer' (bringing
similar surface tension fluids together to make larger and larger sized
particles) and such particles will settle out into a 'drop-out-pot'
..... or usually into the bottom of the filter bowl (bowl pointing
downwards). The smaller the retention size of the filter media the
more efficient the coalescing efficiency.
On the very bottom of the filter bowl, add a pigtail of oil compatible
transluscent plastic tube (Tygon, etc.) with a cock on the end ....
when you see water in the pigtail simply drain the bowl. Thats the
same way a racor with an integral clear plastic knock-out-pot works.

If you are regularly polishing the fuel the coalesced emulsions will be
removed/trapped in the inverted filter bowl .... that why you put the
dip tube for the recirc system at the VERY bottom of the tank and a
drain cock on the inverted filter bowl.

Dont want air leaks or fuel oil leaks ------ dont use compression
fittings, use flared or better fittings.

Pump should have a SCREEN (preferably integral) for protection to
prevent damage by *huge* particles that would tear the rubber impeller
or wobble plate.

I say again, if you want long service life and efficient filtration
employ PRESSURE filtration, especially on a recirculation system.

The ONLY reason I can think of why fuel systems in boats use vacuum
filtration is ........... the engine manufactures supply the lift pump
and 'guard' filter - and puts it on the engine ...... and not on the
tank (where it SHOULD BE). Cheap and dirty solution, easier for the
boat builder - less wiring, less design, less effort, ....

All this begs the question, why does the filter media care whether it's
in "pressure" mode or "vacuum" mode? Sure, the plumbing and filter
cases care. But the media only sees a pressure differential across it.
What's the difference to the media if the there is 14psi (atmospheric
pressure) on one side and, say, 10 psi (a 4 psi vacuum drawing fuel
across the media) on the other side vs. 18 psi (4 psi pressure pushing
fuel across the media) on one side and 14 psi (atmospheric) on the
other?

IOW, even if the pump is past the filter drawing fuel through it, the
filter is still in "pressure" mode because it's really the atmospheric
pressure pushing fuel through the filter.

Steve

On Thu, 08 Jan 2004 14:53:44 GMT, Rick wrote:

Steven Shelikoff wrote:

Well, now I have to ask why the pump should be before the finer stages
of filtration.

Because you can install canned filters which are rated for
high pressure downstream of the pump, on the pressure side.
The filters on the suction side are fairly coarse, produce
little resistance to flow until clogged, and can perform the
initial separation of water and solids.

wouldn't matter where the pump is. Of course if it can't do that (not
enough suction for all the stages before it) then you'd have to move the
pump up in the stream like you suggest above.

If the pump was upstream of all filters, so that it sucked
through them all, the differential available is pretty low.

That all depends on how many "all" is, the max pressure differential you
want to operate at and how much vacuum the pump can draw. 2 stages
should be fine. 3, probably not.

That is why most installations use Racors on the suction
side for the reasons I have given and then the final filters
are can filters on the pressure side.

I have 3 stages with an electric pump between the 2nd and 3rd and then
the engine lift pump after the 3rd. Works pretty good. The only time
it stopped working, leading me to thinking there was a vacuum leak, was
when the tank vent clogged and vacuum built up in the tank to the point
where fuel couldn't be drawn out anymore. Problem solved temporarily by
opening the filler.

Steve

On Thu, 08 Jan 2004 14:53:44 GMT, Rick wrote:

Steven Shelikoff wrote:

Well, now I have to ask why the pump should be before the finer stages
of filtration.

Because you can install canned filters which are rated for
high pressure downstream of the pump, on the pressure side.
The filters on the suction side are fairly coarse, produce
little resistance to flow until clogged, and can perform the
initial separation of water and solids.

wouldn't matter where the pump is. Of course if it can't do that (not
enough suction for all the stages before it) then you'd have to move the
pump up in the stream like you suggest above.

If the pump was upstream of all filters, so that it sucked
through them all, the differential available is pretty low.

That all depends on how many "all" is, the max pressure differential you
want to operate at and how much vacuum the pump can draw. 2 stages
should be fine. 3, probably not.

That is why most installations use Racors on the suction
side for the reasons I have given and then the final filters
are can filters on the pressure side.

I have 3 stages with an electric pump between the 2nd and 3rd and then
the engine lift pump after the 3rd. Works pretty good. The only time
it stopped working, leading me to thinking there was a vacuum leak, was
when the tank vent clogged and vacuum built up in the tank to the point
where fuel couldn't be drawn out anymore. Problem solved temporarily by
opening the filler.

Steve

On Thu, 08 Jan 2004 14:24:02 GMT, Rich Hampel
wrote:

After 30 years of screwing around with this stuff I cant still give a
an accurate technical reason .... my 'opinion' is the 'regime' of
particle depositionIn and the formation of 'filter cake' ..... on a
pressure filtration the deposition begins mostly on the upper surface
or at least within 5% depth of the surface, while with vacuum
filtration the deposition is essentially INSIDE the matrix of the
media. Being inside the matrix causes higher internal velocities
which drive the particles deeper and deeper into the matrix ... causing
an exponential decrease in service life. ..... its the same for depth
as well as membrane filtration.
The quandy is that the fluids are incompressible and shouldnt make any
difference due to the direction of motive pressure .... but in practice
it does, it always does.

Not only shouldn't it make a difference due to the direction of motive
pressure, but the direction of motive pressure is the same in both
cases. Higher pressure on the inlet, lower on the outlet with the same
differential as well. If there really is a difference, then there must
be some other mechanism at work other than just whether the pump is
pushing or pulling.

Steve
article , Steven Shelikoff
wrote:

On Thu, 08 Jan 2004 02:51:46 GMT, Rich Hampel
wrote:

NOPE!
In pressure mode, the filter will also act as a 'coalescer' (bringing
similar surface tension fluids together to make larger and larger sized
particles) and such particles will settle out into a 'drop-out-pot'
..... or usually into the bottom of the filter bowl (bowl pointing
downwards). The smaller the retention size of the filter media the
more efficient the coalescing efficiency.
On the very bottom of the filter bowl, add a pigtail of oil compatible
transluscent plastic tube (Tygon, etc.) with a cock on the end ....
when you see water in the pigtail simply drain the bowl. Thats the
same way a racor with an integral clear plastic knock-out-pot works.

If you are regularly polishing the fuel the coalesced emulsions will be
removed/trapped in the inverted filter bowl .... that why you put the
dip tube for the recirc system at the VERY bottom of the tank and a
drain cock on the inverted filter bowl.

Dont want air leaks or fuel oil leaks ------ dont use compression
fittings, use flared or better fittings.

Pump should have a SCREEN (preferably integral) for protection to
prevent damage by *huge* particles that would tear the rubber impeller
or wobble plate.

I say again, if you want long service life and efficient filtration
employ PRESSURE filtration, especially on a recirculation system.

The ONLY reason I can think of why fuel systems in boats use vacuum
filtration is ........... the engine manufactures supply the lift pump
and 'guard' filter - and puts it on the engine ...... and not on the
tank (where it SHOULD BE). Cheap and dirty solution, easier for the
boat builder - less wiring, less design, less effort, ....

All this begs the question, why does the filter media care whether it's
in "pressure" mode or "vacuum" mode? Sure, the plumbing and filter
cases care. But the media only sees a pressure differential across it.
What's the difference to the media if the there is 14psi (atmospheric
pressure) on one side and, say, 10 psi (a 4 psi vacuum drawing fuel
across the media) on the other side vs. 18 psi (4 psi pressure pushing
fuel across the media) on one side and 14 psi (atmospheric) on the
other?

IOW, even if the pump is past the filter drawing fuel through it, the
filter is still in "pressure" mode because it's really the atmospheric
pressure pushing fuel through the filter.

Steve

On Thu, 08 Jan 2004 14:24:02 GMT, Rich Hampel
wrote:

After 30 years of screwing around with this stuff I cant still give a
an accurate technical reason .... my 'opinion' is the 'regime' of
particle depositionIn and the formation of 'filter cake' ..... on a
pressure filtration the deposition begins mostly on the upper surface
or at least within 5% depth of the surface, while with vacuum
filtration the deposition is essentially INSIDE the matrix of the
media. Being inside the matrix causes higher internal velocities
which drive the particles deeper and deeper into the matrix ... causing
an exponential decrease in service life. ..... its the same for depth
as well as membrane filtration.
The quandy is that the fluids are incompressible and shouldnt make any
difference due to the direction of motive pressure .... but in practice
it does, it always does.

Not only shouldn't it make a difference due to the direction of motive
pressure, but the direction of motive pressure is the same in both
cases. Higher pressure on the inlet, lower on the outlet with the same
differential as well. If there really is a difference, then there must
be some other mechanism at work other than just whether the pump is
pushing or pulling.

Steve
article , Steven Shelikoff
wrote:

On Thu, 08 Jan 2004 02:51:46 GMT, Rich Hampel
wrote:

NOPE!
In pressure mode, the filter will also act as a 'coalescer' (bringing
similar surface tension fluids together to make larger and larger sized
particles) and such particles will settle out into a 'drop-out-pot'
..... or usually into the bottom of the filter bowl (bowl pointing
downwards). The smaller the retention size of the filter media the
more efficient the coalescing efficiency.
On the very bottom of the filter bowl, add a pigtail of oil compatible
transluscent plastic tube (Tygon, etc.) with a cock on the end ....
when you see water in the pigtail simply drain the bowl. Thats the
same way a racor with an integral clear plastic knock-out-pot works.

If you are regularly polishing the fuel the coalesced emulsions will be
removed/trapped in the inverted filter bowl .... that why you put the
dip tube for the recirc system at the VERY bottom of the tank and a
drain cock on the inverted filter bowl.

Dont want air leaks or fuel oil leaks ------ dont use compression
fittings, use flared or better fittings.

Pump should have a SCREEN (preferably integral) for protection to
prevent damage by *huge* particles that would tear the rubber impeller
or wobble plate.

I say again, if you want long service life and efficient filtration
employ PRESSURE filtration, especially on a recirculation system.

The ONLY reason I can think of why fuel systems in boats use vacuum
filtration is ........... the engine manufactures supply the lift pump
and 'guard' filter - and puts it on the engine ...... and not on the
tank (where it SHOULD BE). Cheap and dirty solution, easier for the
boat builder - less wiring, less design, less effort, ....

All this begs the question, why does the filter media care whether it's
in "pressure" mode or "vacuum" mode? Sure, the plumbing and filter
cases care. But the media only sees a pressure differential across it.
What's the difference to the media if the there is 14psi (atmospheric
pressure) on one side and, say, 10 psi (a 4 psi vacuum drawing fuel
across the media) on the other side vs. 18 psi (4 psi pressure pushing
fuel across the media) on one side and 14 psi (atmospheric) on the
other?

IOW, even if the pump is past the filter drawing fuel through it, the
filter is still in "pressure" mode because it's really the atmospheric
pressure pushing fuel through the filter.

Steve

On Thu, 08 Jan 2004 05:54:34 GMT, (Steven
Shelikoff) wrote:

On Thu, 08 Jan 2004 02:51:46 GMT, Rich Hampel
wrote:

NOPE!
In pressure mode, the filter will also act as a 'coalescer' (bringing
similar surface tension fluids together to make larger and larger sized
particles) and such particles will settle out into a 'drop-out-pot'
..... or usually into the bottom of the filter bowl (bowl pointing
downwards). ///

All this begs the question, why does the filter media care whether it's
in "pressure" mode or "vacuum" mode? Sure, the plumbing and filter
cases care. But the media only sees a pressure differential across it.
What's the difference to the media if the there is 14psi (atmospheric
pressure) on one side and, say, 10 psi (a 4 psi vacuum drawing fuel
across the media) on the other side vs. 18 psi (4 psi pressure pushing
fuel across the media) on one side and 14 psi (atmospheric) on the
other?

IOW, even if the pump is past the filter drawing fuel through it, the
filter is still in "pressure" mode because it's really the atmospheric
pressure pushing fuel through the filter.

Steve


Looks like the contribution that mentioned a pressure pump's tendency
to mix and chop big water drops to a clogging emulsion where the
vacuum pump sucks them into the filter intact - that idea didn't
appeal to you?

Brian W

On Thu, 08 Jan 2004 05:54:34 GMT, (Steven
Shelikoff) wrote:

On Thu, 08 Jan 2004 02:51:46 GMT, Rich Hampel
wrote:

NOPE!
In pressure mode, the filter will also act as a 'coalescer' (bringing
similar surface tension fluids together to make larger and larger sized
particles) and such particles will settle out into a 'drop-out-pot'
..... or usually into the bottom of the filter bowl (bowl pointing
downwards). ///

All this begs the question, why does the filter media care whether it's
in "pressure" mode or "vacuum" mode? Sure, the plumbing and filter
cases care. But the media only sees a pressure differential across it.
What's the difference to the media if the there is 14psi (atmospheric
pressure) on one side and, say, 10 psi (a 4 psi vacuum drawing fuel
across the media) on the other side vs. 18 psi (4 psi pressure pushing
fuel across the media) on one side and 14 psi (atmospheric) on the
other?

IOW, even if the pump is past the filter drawing fuel through it, the
filter is still in "pressure" mode because it's really the atmospheric
pressure pushing fuel through the filter.

Steve


Looks like the contribution that mentioned a pressure pump's tendency
to mix and chop big water drops to a clogging emulsion where the
vacuum pump sucks them into the filter intact - that idea didn't
appeal to you?

Brian W

On 08 Jan 2004 11:23:20 +0100, Ole-Hjalmar Kristensen
wrote:

"R" == Rick writes:
///

R This is all assuming you don't have access to a centrifuge which is
R really the best way to handle the process.


R Rick

Do you happen to know if there is any centrifugal separator on the
market suited to small boats? I cannot really think of any technical
reasons why not, but perhaps the market isn't there. Btw., I once had
a Scania truck diesel engine which had a centrifuge for its
lubrication oil. Judging from the amount of gunk it separated out of
the oil, it worked very well.


Now *that* is an interesting question!
The pneumatic tube freezer/heater comes to mind immediately.

I can't recall the appropriate name - but the principle is incredibly
simple: introduce the pressurized fluid tangentially to a short
cylinder, and at one end of the cylinder, the spinning fluid meets a
washer with a central hole - the lighter fraction goes through here.
At the other end of the cylinder, the spinning fluid meets a central
barrier, with an annular gap - so the heavier fraction goes through
here. You can make them with plumbing fittings and minimal
machining or filing.
With pressurized air, these gadgets separate cooled air from heated
air (though the power efficiency is not compretitive with
regular fridges...)

With pressurized fuel, this gadget ought to do a very creditable job
of spinning out water and particles....

Brian Whatcott Altus OK

On 08 Jan 2004 11:23:20 +0100, Ole-Hjalmar Kristensen
wrote:

"R" == Rick writes:
///

R This is all assuming you don't have access to a centrifuge which is
R really the best way to handle the process.


R Rick

Do you happen to know if there is any centrifugal separator on the
market suited to small boats? I cannot really think of any technical
reasons why not, but perhaps the market isn't there. Btw., I once had
a Scania truck diesel engine which had a centrifuge for its
lubrication oil. Judging from the amount of gunk it separated out of
the oil, it worked very well.


Now *that* is an interesting question!
The pneumatic tube freezer/heater comes to mind immediately.

I can't recall the appropriate name - but the principle is incredibly
simple: introduce the pressurized fluid tangentially to a short
cylinder, and at one end of the cylinder, the spinning fluid meets a
washer with a central hole - the lighter fraction goes through here.
At the other end of the cylinder, the spinning fluid meets a central
barrier, with an annular gap - so the heavier fraction goes through
here. You can make them with plumbing fittings and minimal
machining or filing.
With pressurized air, these gadgets separate cooled air from heated
air (though the power efficiency is not compretitive with
regular fridges...)

With pressurized fuel, this gadget ought to do a very creditable job
of spinning out water and particles....

Brian Whatcott Altus OK

On Fri, 09 Jan 2004 00:59:52 GMT, Brian Whatcott
wrote:

On Thu, 08 Jan 2004 05:54:34 GMT, (Steven
Shelikoff) wrote:

On Thu, 08 Jan 2004 02:51:46 GMT, Rich Hampel
wrote:

NOPE!
In pressure mode, the filter will also act as a 'coalescer' (bringing
similar surface tension fluids together to make larger and larger sized
particles) and such particles will settle out into a 'drop-out-pot'
..... or usually into the bottom of the filter bowl (bowl pointing
downwards). ///

All this begs the question, why does the filter media care whether it's
in "pressure" mode or "vacuum" mode? Sure, the plumbing and filter
cases care. But the media only sees a pressure differential across it.
What's the difference to the media if the there is 14psi (atmospheric
pressure) on one side and, say, 10 psi (a 4 psi vacuum drawing fuel
across the media) on the other side vs. 18 psi (4 psi pressure pushing
fuel across the media) on one side and 14 psi (atmospheric) on the
other?

IOW, even if the pump is past the filter drawing fuel through it, the
filter is still in "pressure" mode because it's really the atmospheric
pressure pushing fuel through the filter.

Steve


Looks like the contribution that mentioned a pressure pump's tendency
to mix and chop big water drops to a clogging emulsion where the
vacuum pump sucks them into the filter intact - that idea didn't
appeal to you?

Sure did. So does the idea that I'd rather have a malfunction that lets
air into the system than one that lets fuel out. But both of those
points leads to the conclusion that you should suck fuel through the
filter rather than push it through.

However, Rich recommends that fuel be pushed through due to some
tendency of the filter media to work better in that mode. Thus my
question, discounting all other aspects at hand such as the pump
emulsifying the fuel before it gets to the filter, why does the filter
media care which side the pump is on? Fuel is being pushed through it
either way, either by the pump with a higher pressure on the inlet or
the atmosphere with a higher pressure on the inlet.

Steve