• We will be performing upgrades on the forums and server over the weekend. The forums may be unavailable multiple times for up to an hour each. Thank you for your patience and understanding as we work to make the forums even better.

VFD on submersible pump in a body of water

kwired

Electron manager
Location
NE Nebraska
Occupation
EC
The constant velocity pump is also known as a constant pressure pump and is using a pressure xducer for feedback. The pump motor that was installed is a single phase and the controller will be for a 3 wire single phase motor which is configured for residential pump control. The motor and drive are seperate devices on this system.

My mistake on this installation was assuming that the ground fault detection on the drive was sufficient to meet rule 26-956 but in discussions with Pentair there is no defined specification at which the ground fault detection occurs - could be 1A or more. Of course this is not acceptable.

At this point I will most likely install system with a GF breaker feeding the standard start/run winding relay and motor. There is a large market for this VFD application if there was a safe and accepted solution.
Is likely safe enough, problem is NEC throws in a requirement kind of out of the blue for GFCI protection for things that haven't yet been evaluated for use on GFCI's, based on just one real world incident and a bunch of speculation about many other similar situations. These items weren't normally required to have GFCI protection in the past and therefore not required to pass same leakage levels for listing requirements.
 

chris_3000

New User
Location
65a2QTSukfFZgJs
Occupation
ELECTRICAL ENGINEER
Has anyone found a workable solution to installing a ground fault breaker on a constant velocity VFD drive for a submersible pump installed in a lake? I need to satisfy the CEC rules and the OEM pump manufacturers have said their pumps will not work with a GF breaker. I have a cable of 240' from the drive to the pump. Cable installation has passed inspection but I am struggling to find a solution that will actually meet the code and not trip the breaker. I have seen anecdotal discussion of a combination of certain drives and breakers that will work. My concern with trying this solution is that it will only work for a short period and then start tripping - at which point the homeowner will replace the GF breaker with regular 2 pole breaker.
I normally use a ground fault relay with zero sequence CT on the output of the VFD. The relay should be suitable for AC/DC conversion devices such as VFD to avoid the nuisance trip. Little Fuse and Bender have such relays. https://www.benderinc.com/products/ground-fault-monitoring-for-grounded-systems/linetraxx-rcma420/. Normal GFCI breakers cannot handle the harmonics from VFD.
 

bgard

Member
Location
NW Indiana
Occupation
Service manager
The constant velocity pump is also known as a constant pressure pump and is using a pressure xducer for feedback. The pump motor that was installed is a single phase and the controller will be for a 3 wire single phase motor which is configured for residential pump control. The motor and drive are seperate devices on this system.

My mistake on this installation was assuming that the ground fault detection on the drive was sufficient to meet rule 26-956 but in discussions with Pentair there is no defined specification at which the ground fault detection occurs - could be 1A or more. Of course this is not acceptable.

At this point I will most likely install system with a GF breaker feeding the standard start/run winding relay and motor. There is a large market for this VFD application if there was a safe and accepted solution.
Submersible pumps are not designed to be used in lake applications, they require constant water flow over the case of the pump motor, not just being in water to keep them cool. If there is a large market for variable speed pumps, it’s because of misinformation, a variable speed pump and controller cost way more than a standard pump, and a standard pump will outlast the variable speed pump by years, install a cycle stop valve on the standard pump and it will last even longer and the motor efficiency is not much different than the variable speed system
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
I normally use a ground fault relay with zero sequence CT on the output of the VFD. The relay should be suitable for AC/DC conversion devices such as VFD to avoid the nuisance trip. Little Fuse and Bender have such relays. https://www.benderinc.com/products/ground-fault-monitoring-for-grounded-systems/linetraxx-rcma420/. Normal GFCI breakers cannot handle the harmonics from VFD.
Not the same.
When the NEC calls for "GFCI", that is a Class A ground fault device that trips at 5mA (nominal, the range is 4-6mA) in 25msec or less as per UL 943 standard. Those Bender devices are for what would be referred to as "Ground Fault Protection for Equipment" or GFPE (also sometimes called GFEP), which in North America is NOT considered adequate for protecting people. This should not be misconstrued because the difference could mean someone's life.
 

kwired

Electron manager
Location
NE Nebraska
Occupation
EC
Submersible pumps are not designed to be used in lake applications, they require constant water flow over the case of the pump motor, not just being in water to keep them cool. If there is a large market for variable speed pumps, it’s because of misinformation, a variable speed pump and controller cost way more than a standard pump, and a standard pump will outlast the variable speed pump by years, install a cycle stop valve on the standard pump and it will last even longer and the motor efficiency is not much different than the variable speed system
True for a pump intended to be dropped into a well casing, there are other types of pumps that are submersible.
 

tortuga

Code Historian
Location
Oregon
Occupation
Electrical Design
Not the same.
When the NEC calls for "GFCI", that is a Class A ground fault device that trips at 5mA (nominal, the range is 4-6mA) in 25msec or less as per UL 943 standard. Those Bender devices are for what would be referred to as "Ground Fault Protection for Equipment" or GFPE (also sometimes called GFEP), which in North America is NOT considered adequate for protecting people. This should not be misconstrued because the difference could mean someone's life.

OP states he needs to meet Canadian Electrical Code (CEC) section 26-956 1) d) and it appears that device does meet the CEC specs:
CEC code rule 26-956 1) d) is applicable in Canada. And it partly reads: "ground fault protection shall be provided to de-energize all normally ungrounded conductors supplying the submersible pump, with a ground fault trip setting adjust to function as low as practicable to permit normal operations of the pump, but in no case shall the gf current setting be >10ma for an operating time period exceeding 2.7s;"
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
OP states he needs to meet Canadian Electrical Code (CEC) section 26-956 1) d) and it appears that device does meet the CEC specs:
Huh, interesting. I wish the NEC would make a rule change to that effect... Pond / fountain pumps with VFDs have become a real issue here. I had a little side business for a while making pond waterfall controls for a local water feature contractor that used a small VFD to vary the flow** in order to mimic more natural settings. The NEC change to require GFCIs basically put an end to that, I couldn't find a combination of VFD + GFCI that worked. The Siemens/Pentair combo did, but nobody wanted to pay the cost of a Pentair pool pump for a waterfall feature.

**Funny side note on this; I used a feature in some VFDs that allows you to program a series of "steps" of speed and time, so that I didn't need a PLC or something to "randomize" the patterns. If I set up 5 steps, people could recognize it as a repeating pattern. 6 steps and some people were fooled, but with 7 steps nobody noticed that it was repeating! I always thought that was a weird quirk of human perception.
 
Top