Undervoltage Causing Fire?

[d0nut]

I am working on a project that has two large 480V/3P chillers that are starting across the line. The utility transformer is not large enough to handle the starting current of the chillers. When these chillers start, the 480V service experiences a voltage drop of 15-20%, depending on the other building loading. The facility was experiencing motor and control board failure that led to the discovery of the undervoltage condition. There also appears to be a controls issue with the chillers as some days they short-cycle and will have 40-80 starts, exacerbating the undervoltage problem.

A few days ago, the controls section of a vertical fan coil unit in the building started on fire. The fan coil unit is a 277V unit, fed from the service experiencing the undervoltage condition. I don?t think that the fire is related to the undervoltage problem that I am correcting, but that is outside my expertise. Has anyone ever seen or heard of an undervoltage condition causing a controls board to fail in such a way that it would start on fire?

 

[SceneryDriver]

I've seen sustained undervoltage - lost phase on a 480v supply feeding a 480v/120v control transformer - kill the coil on a 120v-coil contactor. I wouldn't see transient undervoltage conditions causing your problems, however. Perhaps there are other power quality problems being caused by the chillers kicking in. Is the POCO switching PF correction caps into the supply, causing problems?



SceneryDriver

 

[broadgage]

With some types of load including induction motors and switched mode power supplies, a drop in voltage will cause an increase in current.
This might in theory be a cause of fire under certain unlikely conditions, but it should be gaurded against by the design of the equipment and installation.

In the case of an induction motor, the extra current shouldbe tolerated for a short time, and if it be prolonged then the motor overload should protect against damage, let alone fire.

Small motors built into appliances etc do not usually have any overload protection (beyond the very limited protection given by the branch circuit breaker) such motors usually have a thermal trip built in. Other small and cheap motors built into equipment might fail if subjected to prolonged under voltage, but the equipment SHOULD be designed such that the motor will fail safely. That is you may need a new appliance, but not a new house !

Processess or machines that are reliant on pumps or fans for cooling might in theory be liable to overheating if the fan or pump slows or stops, but again if the process or machine is properly designed this should not result in danger.

Some gas or oil burning appliances are liable to fault or failure, but if this causes fire rather than simple plant stoppage then it indicates defective design.
In a gas furnace for example, the contactor controlling the fan might drop out at 65% voltage, but the burner might still operate down to say 40% voltage. That could in theory be a fire risk, but in practice should be detected and prevented by other means, such as a limit stat and an airflow switch that proves air flow before the gas valve can be opened.

 

[d0nut]

I doubt the POCO even knows the chillers are starting. The starting current of the chillers is approximately 2000A. The utility transformer is a 750 kVA transformer, which is what is causing the voltage to drop. It is not a sustained undervoltage condition. It only seems to occur when the chillers are starting. The utility primary is a 13.2 kV system, so I imagine the chillers would have very little impact on that system.

The fan coil that started on fire only had heat/fire damage in the controls section. The motor and electric heat coil only had smoke damage. I think that the cause of the fire was unrelated to the undervoltage issue, but I don't want to give the owner any incorrect information.

 

[Jraef]

I doubt the POCO even knows the chillers are starting. The starting current of the chillers is approximately 2000A. The utility transformer is a 750 kVA transformer, which is what is causing the voltage to drop. It is not a sustained undervoltage condition. It only seems to occur when the chillers are starting. The utility primary is a 13.2 kV system, so I imagine the chillers would have very little impact on that system....

If the chillers are on a separate transformer than the rest of the building, then I would agree that the effect of starting voltage drop may not likely affect the primary side and get down to other loads fed from other transformers, although we don't really know the capacity of the system, even though it is 13.2kV.

But if the 750kVA transformer is feeding the ENTIRE building, including the chillers, then transient or not a 15-20% voltage drop 40-80 times a day is likely to cause incremental damage to other equipment. If the fan was controlled by a PC board controller, and that is what fried, I would not be at all surprised. That control system likely had an input voltage tolerance range and most of the time it will be something like +-10%, maybe 15%, but rarely does anyone design for a -20% voltage drop, because that is supposed to be prevented from happening. That's also the typical "drop out" voltage on a standard AC coil for contactors and relays, so hovering around there can cause chattering, which fries things really fast.

 

[kingpb]

Has nothing to do with problem, just curious, is that 2000A for both starting at same time, or is that 2000A a piece? What size chillers are they?

Regardless, that large of a VD on starting was negligence on the designers part. We love to fix these types of problems, fun to do and consulting fees are always good.

 

[electric_cal]

I am working on a project that has two large 480V/3P chillers that are starting across the line. The utility transformer is not large enough to handle the starting current of the chillers. When these chillers start, the 480V service experiences a voltage drop of 15-20%, depending on the other building loading. The facility was experiencing motor and control board failure that led to the discovery of the undervoltage condition. There also appears to be a controls issue with the chillers as some days they short-cycle and will have 40-80 starts, exacerbating the undervoltage problem.

A few days ago, the controls section of a vertical fan coil unit in the building started on fire. The fan coil unit is a 277V unit, fed from the service experiencing the undervoltage condition. I don?t think that the fire is related to the undervoltage problem that I am correcting, but that is outside my expertise. Has anyone ever seen or heard of an undervoltage condition causing a controls board to fail in such a way that it would start on fire?

Looks like you have multiple problems here. The excessive starting and stopping of the chillers 40 to 80 times a day is not good and could lead to premature failure of the motors if not corrected. I wouldn't think the system controls were designed to operate this way.

The voltage sag issue when starting the chillers is another matter. Can you provide the chiller motor HP? What is the duration of the sag? Installing soft-starts on the chillers to help alleviate the across the line current draw and voltage sag during the starting of the chillers is an option, but only if you can resolve the excessive starting and stopping of the chillers.

Have you contacted the utility and requested that they do a power study on the system?

 

[d0nut]

we don't really know the capacity of the system, even though it is 13.2kV.

The 13.2 kV system pretty new, and has been installed in an area expecting a lot of development that hasn't happened yet. I would imagine the utility system was designed with this expansion in mind and is very lightly loaded, but I do not know for sure. There are other services to the building as well as this one. This service has mainly mechanical equipment on it.

Has nothing to do with problem, just curious, is that 2000A for both starting at same time, or is that 2000A a piece? What size chillers are they?

Regardless, that large of a VD on starting was negligence on the designers part. We love to fix these types of problems, fun to do and consulting fees are always good.

The 2000A is for one chiller starting, but they do not start at the same time. If I remember correctly, they are 260 ton units. If I was the one originally designing the building, there are several things that I would have done differently...

Looks like you have multiple problems here. The excessive starting and stopping of the chillers 40 to 80 times a day is not good and could lead to premature failure of the motors if not corrected. I wouldn't think the system controls were designed to operate this way.

The voltage sag issue when starting the chillers is another matter. Can you provide the chiller motor HP? What is the duration of the sag? Installing soft-starts on the chillers to help alleviate the across the line current draw and voltage sag during the starting of the chillers is an option, but only if you can resolve the excessive starting and stopping of the chillers.

Have you contacted the utility and requested that they do a power study on the system?

Yes, the chiller short cycling and voltage drop are two separate problems. I am working with a mechanical engineer to fix the cycling problem, and plan to fix the voltage drop by going to a wye-delta starter which will reduce the starting current to approximately 700A. I think I will suggest metering the service after the starting method is changed before suggesting replacing the utility transformer with a larger size.

The owner had originally contacted the utility company because the facility was experiencing an above average number of motor and controller failures, especially for a building as new as this one is. The utility company did a power study, told the owner that the utility system was fine, and then the owner hired us.

When we were doing the study, the owner said that a fan coil had started on fire, and asked if it was caused by the undervoltage problem. I think it was just an unfortunate coincidence, but I don't want to tell the owner to replace the other 300-400 fan coils in the facility if I don't have to. And I especially don't want to tell him that everything is fine with the fan coils if it is not.

 

[kwired]

Many utilities do not want to see across the line starting of loads that large, unless maybe you want to pay more for equipment that can deliver the necessary power for across the line starting.

Also kind of hard to believe that a variable speed drive wouldn't be desired for optimizing performance anyway.

Just my $.02.