Pump redundancy - backup

karimwassef

karimwassef

Active member
I travel a lot so I need automated self-healing systems wherever possible.

I've addressed most of the potential failure modes.

I run an open flow system with little to no mechanical filters.

Now I'm at pump redundancy. This is to address main circulation pump failure due to:

1. Calcium buildup
2. Intake clog/impeller jam
3. Electrical/mechanical/thermal failure
4. Outlet clog

It doesn't address a power outage.

So my initial idea was to run a backup pump with a one way valve (yes-hate these) into the main return line. If the output flow drops, I turn the backup pump on until I'm back from my trip.

Unfortunately, if the primary pump is dead, it would short circuit the flow. If the main output is clogged, this wouldn't help either.

So, the only option is to run a separate return line up to the DT. That requires a lot more plumbing and distribution as well as a higher siphon break so that it doesn't act like a siphon under normal (primary pump) operation.

Any better ways to do this?
 
I have wrestled with this for years. The main pump and heater are the two pieces of equipment that would be most problematic in the event of failure. Heater redundancy is pretty easy to manage. I just have an extra one in my sump that I can press into service should the need arise (controlled via my apex). Pump redundancy is not so easy. Though I have thought about what you propose, I have not done it. Probably that is the best approach if one is going to have 'seamless switching'. Not without issue, if course, including the things you noted (plus I really don't want a further return cluttering my system). I do keep a drop in spare pump, complete with preinstalled valve couplings, so that my kids could easily swap it in while I'm away. Not useful if there's nobody to do it though :). My approach, frankly, has been to employ the most reliable pump I can. Since using my personal favorites, I have not had a main pump failure in over 25 years. Periodic maintenance is important, and the flow sensors to shortly be available may prove to be an indispensable component of 'maintenance', as is the use of screens on the suction end.
 
I have two identical return pumps running at the same time. Each pump is connected to its own return line. When I unplug one pump, the system continue working. The water rise some in the sump, the ATO peeps due to high water level, and loud gruggle sound comes from the siphon pipe. Both return pipes have quick-connect elbow installed at the top for siphon break. Both have been working for 10 weeks without any salt creeps.

Another option is if you can measure electricity draw from one pump. If one pump stops drawing electricity, the other pump turns on. Sensor failure could be catastrophic though.
 
I use redundant sensors and probes too.

I wish I had the option for someone to help while I travel but with my wife's focused animosity of my tank, that's not an option. I once called a buddy to help while I was in China and it might as well have been the end of the world.

So it's got to be automated.

Since I prefer DC pumps for control, noise and power, I can't afford the silly prices on the 'most reliable' types. I just don't get spending thousands on a 3000 gph unless it comes with a million dollar insurance policy in case it ever fails - for any reason - including user error :D.

So I need redundancy.

Multiple return lines in continuos operation is a good idea. I had originally wanted to create an alternating return to either DT end and slowly ramp up/down. Then I figured out a way to make an alternating diverter with PVC and a motor... Saved pump cost, electricity, etc...

With two pumps, you need separate siphon breaks and separate plumbing too. So whether to run one or two continuously is a matter of preference.

I have a surge too and the surge pump kicks in at different times anyway, so the sump water level changes and there are multiple sensors that report in to keep things smooth.

So my concept looks logical, feasible and as simple as it gets?
 
The problem with one-way valve and a sensor is reliability. If one fails, and the second pump turned on, your flow rate will double and may lead to flood. Of course, you can add a float sensor that shut off one pump if the water level rise in the tank, but you will end up with too many points of failures. I personally like simple clean design with the least amount of dependent links.
 
Fair enough. I don't pump anywhere close to 3,000 gph with my system (which is a bit smaller than yours). Not intending a debate on the pros and cons of different pumps, but I have found that a 'reliable' option often pays back in ways not easily (or immediately) measured. I also ran dual pumps in parallel back in the 1990's for almost a decade. After not having either pump fail, I went back to just one the next time around. One failing plays havoc with things like siphon drains, in-sump skimmers, and the like; but would not be disaterous like a sump being off line for days.
 
BalckTip said:
The problem with one-way valve and a sensor is reliability. If one fails, and the second pump turned on, your flow rate will double and may lead to flood. Of course, you can add a float sensor that shut off one pump if the water level rise in the tank, but you will end up with too many points of failures. I personally like simple clean design with the least amount of dependent links.
Click to expand...

Because I have a surge, I already have a DT level sensor and two in the sump - one refill, one warning.

I use a lot of redundant sensors.

I also have a camera so I can see the flow. The one in the sump will be my indicator to turn on the second pump. It's semi-automated, but safer.
 
ca1ore said:
Fair enough. I don't pump anywhere close to 3,000 gph with my system (which is a bit smaller than yours). Not intending a debate on the pros and cons of different pumps, but I have found that a 'reliable' option often pays back in ways not easily (or immediately) measured. I also ran dual pumps in parallel back in the 1990's for almost a decade. After not having either pump fail, I went back to just one the next time around. One failing plays havoc with things like siphon drains, in-sump skimmers, and the like; but would not be disaterous like a sump being off line for days.
Click to expand...

I need to pump more like 4000gph. I'll go to 2000gph is emergency mode.

Reliability is the thing that doesn't add value until you need it, then it's the most important thing! I turned on all powerheads to keep things alive.
 
Couple options I can think of.

First you need your redundant pump on a solenoid on the outlet, effectively acting like an NRV until opened. You could double your redundancy on these solenoids by having 2 parallel outlets leading to one common outlet line, so if one solenoid fails the other will still activate.

The conditions of opening the solenoid could be:

1) Flow sensor on the main pump going low (or on the drain?)

2) Amp meter on the main pump

3) A couple of inverted high-level float switches in the return pump section (starts up when sump starts to flood).

Any of these can also fail, particularly if they are never used (due to never needing them) or tested, but the more safeguards you have in place the lower the risk.
 
Thanks.

I have a passive secondary sump so it should not flood unless the DT siphon breaks all fail.

I have a high level float in the sump just in case.

I'm planning on using my Apex to control the pumps.

I guess I don't trust a fully automated backup yet. I'd rather use the camera to tell if something is wrong and then manually control the pumps.

It's certainly better than no backup at all.
 
Interested to see what you come up with.

Thing that would worry me about a backup is whether it would work when you needed it to, and whether water trapped in the piping would stagnate. I think you'd have to run it periodically - rather like those whole house generators that run weekly to ensure they will work when you need them to work. I run my backup ATO pump briefly each day to ensure that it is functional.
 
Assuming the pump, and part of the piping, will be below the level of the sump (or in the sump), there will be some. Whether it's material or not, I cannot say.
 
Neptune has an interesting possibility in the pipeline (ba dum tiss) with the FMM. They introduced a flow monitoring system for the Apex that will allow us to monitor flow in our plumbing. One could plumb the return pumps separately and program the second one to run of flow from the first drops.

Looking back, you mention 3000 GPH. You know, a pair of Ecotech Vectra M1 might be in consideration. Their combined flow could be dialed to around 3000, an individual failure would still allow the other to run an acceptable amount of flow for life support, and you would be able to remotely monitor the status. Cost: about $700.
 
Im running a Jebao DCT15000 that's at 3000gph at 4ft, 4000gph at zero head. It hasn't failed me yet, but I have a DCT12000 backup I want to plumb in - just in case.
 
Breadman03 said:
Neptune has an interesting possibility in the pipeline (ba dum tiss) with the FMM. They introduced a flow monitoring system for the Apex that will allow us to monitor flow in our plumbing. One could plumb the return pumps separately and program the second one to run of flow from the first drops.

Looking back, you mention 3000 GPH. You know, a pair of Ecotech Vectra M1 might be in consideration. Their combined flow could be dialed to around 3000, an individual failure would still allow the other to run an acceptable amount of flow for life support, and you would be able to remotely monitor the status. Cost: about $700.
Click to expand...

This is exactly what I recommended above, except you don't need to wait for Apex since flow meters already exist.
 
I think a flow meter makes sense, but only to cause an alarm to go off so I can visually inspect via camera. The switch over decision I would still prefer to do myself.
 
I have two level switches.

LOW = need to add RODI water
HIGH = about to overflow - something is wrong.

If all the pumps are off, the HIGH should still not be triggered. It only gets triggered if there's an imbalance of some kind. ATO malfunction, Kalk doser failure, etc...

It's my catch all in case a bunch of other sensors fail and the sump is about to overflow.
 
You must log in or register to reply here.

Similar threads

S
Sump at Same Level as DT
Replies
9
Views
523
wvned
wvned
C
Help with aio return pump.
Replies
6
Views
474
Cruzersmith
C
C
Well pump and R/O system
Replies
12
Views
893
zappa
Z
Anemone
Power Outages
Replies
3
Views
447
alton
alton
E
Wave maker recommendations?
Replies
12
Views
665
ExitTrauma
E
Back
Top