MRI machine audible at outside 300kva transformer

[JimboSlice]

Hello. First post for me.
I have a question about a situation that is a little over my head as a service electrician. I recently was called to an imagining facility that does MRI scans. They had a MRI tech evaluate the system before they install a new MRI machine. In addition to loose connections in various switchgear and panels that i have already addressed, on of the issues he described is "the 300kVA transformer is maxed out. As you stand by it, outside, you can hear the transformer struggle with the loads-you can hear each pule of the MRI. Not normal. When new equipment is planned, ensure the power is upgraded to take into account new loads, and to provide some overhead for the transformer."
The transformer is a 300kva 3 phase 208/480 step up transformer, that feeds a 480v panel with 3 125 amp breakers, that feed the MRI machine, and 2 other pieces of equipment. I put a data logger on the primary side of the transformer, on the load side of a 600 amp breaker that is the primary OCD of the transformer. I just analyzed the data, and it doesn't appear to be near the capacity of the transformer. The maximum amperage on the primary side is between 350 and 420 amps on the various phases. at inrush peaks. the average amperage is more like 60-70 amps. My question is, is this a "normal" phenomenon, caused by collapsing magnetic fields? Does anyone have experience dealing with these machines? Also, the lights just outside the MRI room where the techs sit, dim significantly at certain high demand parts of the MRI process. however, these lights are on a "house" panel, not part of the switch gear that pertains to this office specifically. And I believe these are the only lights that dim...the ones close to the MRI room itself. Any insight from the super-geniuses would be great.
Thanks

 

[junkhound]

How close are the 300 kVA and the 'house' transformer to the MRI machine ?

There is a possibility that the stray magnetic fields of the MRI machine bias the transformer cores, which causes both higher magnetizing current peaks and magnetostriction noise.

Put a scope with current xfmr on the primary and see if the waveforms are at all distorted - e.g one side of the current waveform will have higher peaks.

My cousin is a retired MD, she recalls years ago when John Hopkins first got a MRI her ferrous earnings were painfully pulled (she was in the next room) when the MRI went into operation.

 

[JoeStillman]

Tell the tech we measure load with an ammeter, not a dBA meter.

The cyclic hum is not unusual. All transformers hum. (they can't remember the words).

P.S. Welcome to the forum!

 

[mike_kilroy]

Tell the tech we measure load with an ammeter, not a dBA meter.

The cyclic hum is not unusual. All transformers hum. (they can't remember the words).

P.S. Welcome to the forum!

Nice reply!

 

[mike_kilroy]

Hello. First post for me.

1) ...he described is "the 300kVA transformer is maxed out. As you stand by it, outside, you can hear the transformer struggle with the loads-you can hear each pule of the MRI. Not normal. When new equipment is planned, ensure the power is upgraded to take into account new loads, and to provide some overhead for the transformer."

2) The transformer is a 300kva 3 phase 208/480 step up transformer, that feeds a 480v panel with 3 125 amp breakers, that feed the MRI machine, and 2 other pieces of equipment. I put a data logger on the primary side of the transformer, on the load side of a 600 amp breaker that is the primary OCD of the transformer. I just analyzed the data, and it doesn't appear to be near the capacity of the transformer. The maximum amperage on the primary side is between 350 and 420 amps on the various phases. at inrush peaks. the average amperage is more like 60-70 amps.

3) My question is, is this a "normal" phenomenon, caused by collapsing magnetic fields? Does anyone have experience dealing with these machines?

I would like to give my 2 cents on 3 things mentioned...

1) IIRC from a little work on MRI machines in Cleveland, OH (we supplied the rotation big frameless motors the person is put inside of, again IIRC from 30-35 years ago), the many many servo amplifiers that powered each magnetic pulse pulled extremely high short length amperage. The result WILL be hearing that in the transformer windings very easily. We also have designed and supplied transformers like this for this and other applications for long time. So for their tech to say that "BIZT" sound is not normal when the MRI 'hits' is just nonsense. His comment to be sure your supply is sufficient to power the present AND new equipment is a nice generic and true statement.

2) This sounds ok on the surface, but running the numbers and it says something is really wrong? Taking it one point at a time:
2a) 300kva @ 208v 3ph input is good for 833 amps continuous, @ 480v 3ph it is rated 361amps. OK, got the specs out of the way...
2b) Know it is typical to put 1.25xrated current breakers on transformers...
2c) You say the panel has 125amp breakers on 480v side?? That is 125*480*1.73= 103kva? So correct size for about a 103/1.25= 80kva (96amp rating) transformer...
2d) You say the panel has 600amp breakers on 208v side?? That is 215kva size, so correct size for about a 175kva (486amp rating) transformer - THIS 600AMP BREAKER WOULD NEVER BE ABLE TO TURN ON A 300KVA TRANSFORMER WITHOUT TRIPPING ON INRUSH 99% OF THE TIME.

So continuing your LOAD numbers:
2e) Your 70a @ 208v is 25kva load. Sure, we all know the inrush at turn on can be easily 10+ times higher momentarily, So I doubt you measured that; I suspect your 420amp is the MRI "hits" as it builds each magnetic field, and I suspect this is the "BIZT" sound the tech noted in the transformer. So that peak is 151kva worth...

So I obviously question your 300va xfmr size. Empirically, I suspect from the breakers, that you really have more like a 175 kva transformer...

MY suggestions are thus:

1) Find the REAL present transformer size so you know what you have today.
2) Call the MRI manufacturer and ask for application engineering; tell them the present transformer size and load measured, and ask THEM to recommend if a larger transformer will be required when you add xytr-244 model new MRI to the circuit. Do NOT let them give you that same tech that came out and made the ridiculous comment.
3) If #2 does not work, at least come back here and post the real transformer size and add the current draw rating of the new MRI being added and folks probably can recommend if a new xfmr is required, and if so, what size THEY would use.

Good luck!

 

[Electric-Light]

You need to log it with a power quality monitor, not an ordinary voltage/amperage data logger.

MRI is a very impulsive load and could require a very stiff source.

Consider a smaller example. Your microwave is designed to operate on a 15A outlet, therefore a transformer that is rated for 2kVA continuous may seem adequate. A desk lamp operating from the same transformer would pulsate like hell as the microwave cycles on and off through the cooking cycle due to the major load step in addition to inrush current. In fact, the control circuit within the microwave may glitch out. You're within the safe thermal limit of the transformer but the high impedance is causing functional issues.

Connect it to PoCo's 40kVA transformer and it wouldn't be a problem.

The MRI scenario you have is the same thing on a bigger scale.

You have:

Poco's primary to 208 transformer supplying 208/120, then on-premise 208 to 480v transformer.
The light dimming in the control room can be from the magnetic field interfering with the ballast's internal parts or the dip in the power source. You could run power from an outlet that does not flicker to the light with an extension cord temporarily to rule out. A data logger that isn't meant for capturing power disturbance wouldn't catch dips.

The utility feeder for a small practice wouldn't have the same stiffness as the loop of a large business park or a hospital campus. You would have to work with the utility if the dip is propagating into the utility feed. A 480/277 utility fed source would be much stiffer.

 

[qcroanoke]

MY suggestions are thus:

1) Find the REAL present transformer size so you know what you have today.
2) Call the MRI manufacturer and ask for application engineering; tell them the present transformer size and load measured, and ask THEM to recommend if a larger transformer will be required when you add xytr-244 model new MRI to the circuit.
3) If #2 does not work, at least come back here and post the real transformer size and add the current draw rating of the new MRI being added and folks probably can recommend if a new xfmr is required, and if so, what size THEY would use.

Good luck!

#1 Agreed
#2 Agreed, and if you get the job you need stamped drawings from a State licensed PE. And follow them exactly!
#3 Agreed, but do what the Manufacturer and PE require you to do.