Roman Aqueduct w/Siphon lock?

jbjb

I realise this is not a straightforward situation and there are lots of ideas here. Just a thought... if the dewatering scheme is faced with a range of demand, are there options to better match the pump capacity to it? For example...

• Variable speed drive to modulate the pump(s)
  
• Multiple pumps of varying capacity, controlled so that overall flow rate increases as water depth in the sump increases. Better load-matching, less short-cycling and all the horrors that brings, redundancy
  

The ideal would be pumping capacity perfectly matching the demand running 24/7 at the lowest possible output. Ultrasonic level sensor over the sump?

farmwi

Enjoying thinking about the possibilities from all the engineers.

Reliability, lower flows, consider using VFD with a pressure transducer to flow match rpm and utilize a 'deep well' pump with a single or fewest impellers and an downsized motor.
I would consider sinking a borehole deeper than the foundation to dewater the foundation. It sounds like there is a large amount of pressure on the exterior foundation walls at high flows. Wouldn't want the building to float. If the base is clear gravel, site it close to the discharge path. If there was no reason against it, I would consider doing it outside the building.

Economically. For larger flows, maybe an easier way to start a siphon with low power for high flow is found in every bathroom. Use a foot valve, a check on the submerged intake, and a storage tank to dump water down the effluent side fast enough to seal the water on two bends, ie make a toilet, start the siphon with a flush. The
VFD could fill a tank with a fill valve for larger the larger flows.
The same VFD with pressure transducer could trigger a flush or a valve operation at different levels or pump speeds.
They usually have two analog pressure inputs and 3 relay outputs. And can handle simple control operations.

3 phase motors pump motors often last 20-30 years with regular cyclical operation.

Jamie Hall

There are so many options! I get a chuckle out of the flush idea -- that's a concept which has been used for at least a century on sewer systems with flat grades. The idea being to have a nice big tank at the upstream end which slowly fills (as you say, like a toiler) and then finally the siphon action starts and empties the tank -- whoosh -- and flushes out the sewer. Same concept for dosing siphons for larger septic leach fields, or ones in bad ground.

One which I haven't seen yet, if the space really is needed, is to use horizontal boring from the basement to run a really big enough pipe (say 8 inch maybe) from the sump straight out to daylight near the canal and forget all the fancy stuff. Not cheap, but I wonder if it really would be that much more expensive overall -- never mind the fact that gravity flow is pretty reliable... I've used that technique where you really don't want to dig a trench (under a railroad track for instance!) but you need to get from here to over yonder somewhere...

RickDelta

"@PaulFormisano"
"@JbJb"
"@Farmwl"
"@JamieHall"

.... All great stuff guys!! Thank-you! : )

I had addressed many of these questions earlier.


Let me recap here as to my current (intended) pumping paradigm:

Remove all electrical and submersible pumps from bottom tier.
Flood bottom basement tier to obtain a very large storage sump basin (1,500 gallons , @18 inches deep max.).
Install four (4) AMT 2876-95 self priming pumps at street level inside building (w/soft-starters).
Install special "shallow" (2ft max. water level) sump pressure level sensor.
Install four (4) magnetic water flow sensors in discharge pipes.
Control everything with touch screen PLC.


Operational flow:

All pumps sleep (level sensor monitoring for nearing breach of allowed water level).
At high level trip, run lead pump (to prime siphonic piping) until proof of established siphon flow , sleep pump, check for siphon flow, if not, resume pump til basin empty.
If water level rises beyond first set trip level, run lead + lag pump one
If water level rises beyond first set trip level and second trip set level, run lead + lag pump one + lag pump two ........ (likewise for lag pump three)

109A_5

Paul Formisano said:

The discharge doesn’t have to be submerged, but the intake does.

My thing with the 'not letting the air back in' at the discharge end is so you would not need the pump to restart the siphon.

If there is a modulating valve inside the building that can stop the flow of water the discharge side will drain out, since the air will get in and you will loose the siphon.

I would eventually build this out (one time cost) to use minimal electricity (an ongoing cost). To me the only electricity needed is for the control, not pumps running, the only time for the pumps is when the ingress exceeds the siphon capability.

Baby steps...

RickDelta

"@109A_5"

Without adding a receiver or any modulating or discharge stop valve ...... my (now revised) thoughts were to simply , at water levels first trip point, turn on all pumps for lets say 60 seconds (to prime pipes), then sleep pumps, check for siphon in both pipes.
if confirmed siphon, just let the siphon continue the discharge work on its own.
if not, resume only lead pump.

For just a 60 second run of all pumps, siphon discharge may last for weeks straight at times without any pumps called in to run.

(with all pumps running .... may only take 30 seconds to prime!)

..... your thoughts?

109A_5

Hello @RickDelta,
Yes the pump run time to prime probably won't take much, then the majority of the flow is through the bypass valve. Maybe experiment to find the minimum reliable prime duration when the discharge pipe is empty. Then add on 20% to 50 % for reliability.

I have no idea what water ingress season you are presently in, dry, mid or overwhelming. I'm hoping the siphon will lower the present water level in a lot less time than weeks.

Presently if the water ingress is not too bad I can see the siphon draining the basement to the point air will break the siphon at the intake. Once that happens you are back into cycling the pumps a lot unless you delay the pumps and let the lowest tier flood again. Personally, eventually, I would try to keep the basement as dry as possible. Maybe this is all just wishful thinking on my part.

That is why I said "Baby Steps" you got to start somewhere and learn as you go, maybe as time goes on refine for the lowest power consumption and best control. I think you mentioned the cycling of the pumps was burning them out. So the way I see it, if eventually if you can use your PLC to manage the system and minimize the pump usage that is what I would do. Less power cost and less pump cost.

To me the cool thing is with the use of the PLC and a water level sensor you could modulate the siphon flow (with some creative valving) to actually keep it in siphon mode almost indefinitely, so the water expelled equals the water ingress and the water is only in the sump pit, all automatic and self adjusting so you don't need to run pump(s) to prime.

Then when the ingress is in the overwhelming season then use the pumps. Since I believe the pumps can push the water at a higher rate than just a siphon.

The other thing is during the drier season when only small siphon flow volume is needed the water may not fill the discharge side of the pipe and since it is quite horizontal, air may make its way in and break the siphon requiring a restart. If a means is implemented so the air can't get into the discharge end the low volume siphon won't need a pump to run (or other means) to re-initiate the siphon.

Jamie Hall

I have no idea -- nor do I want to know at this point in the thread -- what the surrounding soil and ground water levels are. I can only relate that I did a municipal well once, some years ago, in sandy soil which quite happily flowed at 2000 gpm, 24/7, in and out of season -- and was only two feet in diameter with only 8 feet of draw...

Good luck.

leonz

More valuable information on electric motors and "Inrush Current" that should be examined.


"2017 National Electrical Code Article 430.52"

"Per Kirchoff's law, the sum of voltages around a closed loop must equal zero. If we a assume a 0.5 Ohm

source impedence and a 10 ampere nominal current on a 480 volt system. The inrush current can result

in a drop of 30 volts to 50 volts. Therefore, voltage at the load would sag to 430 volts down from the

nominal 475 volt level".

RickDelta

Self Priming Pumps Install

109A_5

Four pumps, two 4" pipes for water removal, does each pump have a check valve for back flow prevention through the other pump on the same 4" pipe ?

leonz

Four 3" pumps for water pumping, pulling water from sump into pump room and then it uses the 3" discharge from each pump on each of the four down pipes from pump room to existing discharge piping in basement.

Each pump case has a rubber check flap between the inlet flange and pump case.

RickDelta

109A_5 said:

Four pumps, two 4" pipes for water removal, does each pump have a check valve for back flow prevention through the other pump on the same 4" pipe ?

yes! .... built-in the pump.

RickDelta

"@PaulFormisano"

If you form a slight horn shape, flared out at the end of the discharge pipe, right after the valve, the flow through the pipe will increase slightly vs. a sharp cut.

....... what is the engineering reason for this increase of flow??

Contained high pressure immediately dropping into a low pressure environment??

Sounds like an aiding form of "reverse" water cavitation.

RickDelta

Question:

I'm assuming I will need to install an "air admittance" or "air bleeder" valve at the highest point in my discharge loop.

leonz

No, you don't need to do that Rick, you need to be able to prime the pump chamber when needed.

If you install a basket strainer at the bottom of the suction pipe for each dewatering pump you will always have water in the riser and the pump chamber as long as you have a swing check valve above it.

I am sending you more information via a PM.

109A_5

RickDelta said:

Question:

I'm assuming I will need to install an "air admittance" or "air bleeder" valve at the highest point in my discharge loop.

For what purpose ? It would only terminate the siphon, by letting air in. The pumped water will easily push the air out of the discharge end of the pipe.

The only thing you would need at the highest point or above the pump(s) is if you use vacuum to prime the pump(s).

neilc

brings up an interesting question,
where is this? ( I'm to lazy to go back and read 4 pages),

Is this subject to freezing?
Can the syphon stay filled? winter wise?

neilc

ok, Delaware,

Is this subject to freezing?

RickDelta

109A_5 said:

RickDelta said:

Question:

I'm assuming I will need to install an "air admittance" or "air bleeder" valve at the highest point in my discharge loop.

For what purpose ? It would only terminate the siphon, by letting air in. The pumped water will easily push the air out of the discharge end of the pipe.

The only thing you would need at the highest point or above the pump(s) is if you use vacuum to prime the pump(s).

"@109A_5"
Isn't it possible for the air to be trapped in the high end loop where the water actually has to "tunnel" its way under the entrapped air reservoir ...... effectively reducing the pipes capacity to flow water?

My discharge pipe goes up 12ft , horizontal for 14ft then straight down for 6ft ...... then begins its long 330ft horizontal run to canal

RickDelta

neilc said:

brings up an interesting question,
where is this? ( I'm to lazy to go back and read 4 pages),

Is this subject to freezing?
Can the syphon stay filled? winter wise?

Bristol, PA
.... never had any pipe freeze ups to the outside discharge pipes.

RickDelta

"@Leonz"

This is the caged full port "foot valve" I was thinking about:
https://sprinklerwarehouse.com/munro-pvc-foot-valve-3-in-fpt-fv300t

neilc

RickDelta said:

neilc said:

brings up an interesting question,
where is this? ( I'm to lazy to go back and read 4 pages),

Is this subject to freezing?
Can the syphon stay filled? winter wise?

Bristol, PA
.... never had any pipe freeze ups to the outside discharge pipes.

but you weren't keeping that pipe full of water trying to hold a syphon,

RickDelta

neilc said:

RickDelta said:

neilc said:

brings up an interesting question,
where is this? ( I'm to lazy to go back and read 4 pages),

Is this subject to freezing?
Can the syphon stay filled? winter wise?

Bristol, PA
.... never had any pipe freeze ups to the outside discharge pipes.

but you weren't keeping that pipe full of water trying to hold a syphon,

Correct! ..... but my thoughts were to let the siphon "break" (no standing water outside) if water level gets that low.
Next pump cycle, run all pumps for 60 seconds to re-prime siphonic discharge pipes.

Foot note:
1/4 of the 330 discharge pipe is in fact exposed to freezing temperatures.
..... you made a good point here!
Thanks! : )

Jamie Hall

Air release valves (automatic) are commonly used on high points on aqueducts, s they assist in filling. There may also be vacuum relief valves at the same locations, to prevent the aqueduct internal pressure falling below atmospheric )but in this application you do want the internal pressure to fall below -- in fact, well below -- so vacuum relief is neither needed nor wanted).

However, in this instance I would expect the velocity in the pipes would be adequate to remove trapped air.

RickDelta

Intresting foot note:

In this entire bottom basement tier i've committed to a large water storage basin (to eliminate the short cycling issues) ...... has a doorway with one step down (landing to landing 18" deep). This works out to aprox, to 1,500 gallon of storage area for this size room.

Now, lets say I brick up the doorway 18 inches ..... thinking to easily double my storage to 3,000 gallons!

This won't work here : (
The water level on tier two would be exactly the same as the bottom tier due to penetrations throughout tier two and even tier one.

Jamie Hall

Somewhat off topic -- but the heading suggests it... the Roman aqueducts really are fascinating to study (never mind look at). Whatever else the Romans might have been, they were marvelous engineers. The lengths they went to to get steady gravity open channel flow are impressive. So far as I am aware, though, they never used siphons -- probably because for all their skill making vacuum tight pipe was not something they could do on a large scale. They did use some pressure pipe dips (sometimes called, quite misleadingly, "inverted siphons"), though, for some very wide valleys where it just wasn't practical or feasible to construct an elevated aqueduct structure. They aren't common at all (leakage problems again) but they did exist, and pressure pipe water supply was common enough in their larger cities (along with what we would regard as primitive -- but remarkably effective sewage works) (Rome still uses some of the Roman sewers in their storm water management system).

Some of the other early water supply schemes are pretty impressive, too, and worth studying -- the Greeks had an interesting arrangement for Athens, for instance, and the qanats which the Persians built in what is now Iran are too (they also took advantage of the insanely dry climate to cool their buildings by evaporation, and even, so it is said, managed refrigeration that way).

RickDelta

"@jamiehall"

Jamie, you wouldn't have lasted very long with all your knowledge back in those days.
You'd walk into Caesars office and say " lets just hook up a few 10hp 3-phase pumps with with schedule 40 pipe distribution and be done with it"

......... you'd be the main event at the coliseum that night!

leonz

RickDelta said:

"@Leonz"

This is the caged full port "foot valve" I was thinking about:
https://sprinklerwarehouse.com/munro-pvc-foot-valve-3-in-fpt-fv300t

---------------------------------------------------------------------------------------


Turbulence would be my major concern with air bubbles.
The metal basket strainers are practically wide open in design with the perforated metal walls of the basket and trapped air would be negligable as it would always be rising above the basket strainer.

RickDelta

"@Leonz"

I don't really need to know , even after the head losses come into play, exactly what the GPM value my pumps are pumping.

Digital vacuum sensor $125.00 (no moving parts inside)
3" Magnetic Flow sensor $1,300.00 (no moving parts inside)

I currently am thinking to use flow sensors but couldn't I simply use a vacuum sensor to my PLC to indicate the normal expected flow rate by its developed vacuum pressure?

Lets say, a plastic bag gets into pump basin and ends up blocking the suction intake pipe. I would immediately know there was a problem with that pump.


..... your thoughts?

Jamie Hall

Good thinking on the plastic bag, because it or something like it will get in there, sooner or later. I'd stick with flow sensors, though. And rig them, maybe with a time delay, to turn off the pump before it cavitates long enough to destroy itself.

leonz

Yes, you would know, and the pump starts screaming and heating up from suction cavitation.

First:

www.youtube.com/watch?v=ZQKpu

Second:

www.youtube.com/watch?v=T8jcNAPZ5is&t=4s


Suction cavitation and discharge cavitation:

ipe-pumps.com/process-equipment/pump/cavitation-cause-effects-and-how to avoid/

------------------------------------------------------------------------------------------

I believe the reason you were losing so many sump pumps was due to discharge cavitation.

If it were me I would use 2 pairs of mercury tilt switches zip tied to a piece of steel pipe placed inside a large diameter perforated PVC pipe with water holes drilled in the side every inch to allow water to enter and exit the pipe which is the exact opposite of using a 5 gallon bucket with holes drilled around it from top to bottom to allow a sump pump to drain a household sump and keep it from being damaged.

You could use black zip ties to tie the steel pipe to the larger one through the holes you drill in the larger pipe to keep it in place.

With enough holes drilled into a large diameter piece of capped PVC pipe with a weight(rocks)in the bottom it will resist floating and if needed you could weight it down with a piece of thick wide steel flat stock too.
The rocks would hold the small pipe in place too.
That way you could let two pumps run and stop and the second set of pumps run and stop with the tilt switches based soley on the water level.

My thoughts anyway

109A_5

If you really want to know the water level in the horizontal pipes at the highest part of the system, you could add a sight glass. Maybe also a 4" tee off the main pipe flow so the flow does not bias the sight glass level reading.

neilc

ice, ice, baby,

RickDelta

Jamie Hall said:

Good thinking on the plastic bag, because it or something like it will get in there, sooner or later. I'd stick with flow sensors, though. And rig them, maybe with a time delay, to turn off the pump before it cavitates long enough to destroy itself.

Well , the thing is the 3" mag flow sensors cost as much as the pumps!
: (

Why not just use the low cost vacuum sensor monitored by the PLC.

Flow reference indication to suction pressure would be easy, when pumps are new and run for the first time, simply note the digital pulse train out of the sensor and visually show this on the pump flow bar graph.

Any "out of range" vacuum pressure would alarm and shut down pump.

As I said, I don't need to know GPM value ...... only if pump is struggling and in harms way.

RickDelta

"@leonz"
I would use 2 pairs of mercury tilt switches

Since my water storage basin is only 18" deep, I have to detect all four level set points all within 1/4" of each other! (these set points relate to pumps 1 thru 4)

I'm using a special "shallow" depth digital diaphragm sensor. Mercury "tilt" float switches would be almost impossible to use for this.

leonz

Desk Reference Page 823 last column

Using 2 pumps with 4 inch basket strainers and 4' flap check valves,

A 50 foot section of schedule 40 3-inch PVC Pipe with an elbow and short piece of pipe in the sump raised slightly to drain into the sump would theoretically allow almost 574 gallons per minute of syphon flow into the sump once the pump or pumps start.

If the tilt switches were placed away from the sump in calm surface water in the basement mounted on a pressure treated 2" by 2" they would be more effective in controlling the pump or pumps and dewatering the basement.

Jamie Hall

If your switches are going to be on that close a spacing, with that size pumps and that size sump, you are going to need to place them in a very good damping chamber or they will switch on and off like crazy.

RickDelta

Jamie Hall said:

If your switches are going to be on that close a spacing, with that size pumps and that size sump, you are going to need to place them in a very good damping chamber or they will switch on and off like crazy.

With the storage surface area being so large, even in times of extreme ingress water flow ......... the surface is smooth as glass and crystal clear ..... you can see everything thru 18" of water to the bottom floor! .
The underwater turbulence is contained within the deeper original sump pit.

At low ingress flow times times, It may take a week to rise 1/4"!
At extremely high ingress flows, It may take 60 seconds to rise 1/4".

The 1/4" set point separations are not even an issue.

Foot note:

" .....a very good damping chamber"

I dampen (cancel) any abrupt "surface" level variations "waves" in software with a Kalman filter.

ie: You step into the storage basin, your foot causing a 3/4" wave.

The "peak" and "trough" cancel out to zero level change!



Foot note:
This is how clear and calm the water level is even with 7 pumps barely keeping up with the water ingress. No turbulence!

Jamie Hall

Ah yes. Surely we can get a nice computer program to fix things up when we get the fundamentals wrong... the modern age.