5 ms·

I work on one of the densest power grids in the world (NYC) on the team which does emergency transformer replacements and can possibly answer questions about th

by ics 7y ago

I work on one of the densest power grids in the world (NYC) on the team which does emergency transformer replacements and can possibly answer questions about this. From failure to in service for a single transformer in Manhattan is down to 2-3 weeks. That includes design and engineering (my part), permitting, and construction. For fast transformer replacements we rely on two paths, either system spares (identical units, sometimes refurbished or like in kind replacements) and more recently mobile resiliency transformers. Resiliency transformers are trucked in on semis and can operate at 69, 138, or 345KV in our case. (There's some lightweight reading on them here: https://www.tdworld.com/overhead-transmission/siemens-introduces-demand-mobile-resilience-transformer-leasing https://www.tdworld.com/overhead-transmission/siemens-introd...)

MisterTea 7y ago
Neat job. I've always wanted to poke my head into ConEd's system and see the whole thing from behind the scenes. Years back I remember ConEd was replacing two or three very large transformers down the street of a large venu I was working an event. Turns out we worked the same event the next year and amazing the same day ConEd again had the second transformer scheduled to moved into place. What an operation. Huge trailer from Bay Crane along with one of their giant hydraulic cranes. Was told the transformer weighed in at 240,000 pounds.
- ics 7y ago
  Heh, our most recent replacement a few weeks ago came in at almost 900,000lbs with oil. Depending on the speed of the replacement you can't always move the transformer dry because you need a certain amount of time between filling it and putting it in service.
  - jacobush 7y ago
    Air bubbles need to rise first?
  - MisterTea 7y ago
    That sounds like quite the transformer. That must have been at one of the big 345kV subs. And I assume shipping filled is to avoid the time needed to remove air bubbles. I know a few people in the ConEd trenches, one guy is a lineman who does all the >=69kV stuff and the other guy worked in substations and was in mott haven last I spoke to him (cushy job). I'm always amazed how much new tech con ed has installed and how much old stuff is still kicking around. I live in south queens and we're all 4kV overhead fed from area subs in turn fed by 27kV. Some of those 4kV poles have westinghouse transformers from the 30/40's still in service, glass insulators, and leftover wooden pins from older insulators. Some are so old the primaries have fallen off and just hang between two adjacent poles, some the crossarms are half falling off. Amazing this grid doesn't go down more. Then the next pole is brand new, composite strain insulators, cutouts, reclosers, metering, etc. All modern hardware. And I know some areas they upgraded the 4kV overhead to 27kV. The secondaries are in even worse condition, ancient cloth insulation half weatherd/baked off. But it all still works. I really love how there's this massive network of cabling and systems which orchestrate the flow of power from multiple synchronized generators so it can get to your humble wall socket and charge a phone or turn a light bulb on.
    - snowwindwaves 7y ago
      They are shipped without oil and full of nitrogen gas to reduce the weight.
ethbro 7y ago
Are transformers like these standardized, or are they generally built as one-off customs for the specific site? (I'm not an EE, so not sure how many variables one needs to contend with in grid-scale transformer design)
- ics 7y ago
  We have purchasing standards which are particular to our system, though specs may match others' by coincidence. Most of these are electrical, e.g. how the primary and secondary windings are configured (delta-wye in most cases) but also what gets mounted to the outside (bushing potential devices, current transformers, gas monitoring units, etc. These are not unique to each site which is why there can be so much work to replace one. Older stations were actually designed around the transformers in some cases and the physical configuration of a 50+ year old failed unit is rarely like those produced more recently. We might have a unit on hand that is electrically equivalent but impossible to install without essentially rebuilding the entire station. To give a rough overview of the range of transformers on our system, the first thing we care about is the voltage and MVA rating, then whether the bushings are open air or gas insulated, what the winding configuration of the primary and secondary are (delta-wye being most common), whether the unit it is replacing has a tertiary winding (sometimes at transmission voltage, other times distribution or for light and power to the station), and then maybe what accessories it comes with such as current transformers for protective relaying.
asdfasgasdgasdg 7y ago
I'm curious how long that process would take if it were do or die. Obviously, permitting would no longer be involved. So the real question is how long is the critical path through engineering->construction?
- ics 7y ago
  It varies significantly from project to project. What we depend on is the accuracy of existing records (30-60 year old microfilm scans typically, schematics and physical drawings), similarity of the failed and replacement units, ease of access to the site, support needed by other disciplines (e.g. concrete or steel work), and cost even if it is do-or-die. A perfect like in kind replacement could possibly be done entirely by construction using proper procedure of documenting their disconnects if there were no structural modifications required to move the units in/out. This is rarely the case however and so we need to check everything from phase clearance on making the high voltage connections to whether disconnecting a failed transformer will trip out another perfectly healthy one, possibly making the situation much worse. All that being said, we make every attempt to triage the work required and get information to construction as soon as possible so that no time is wasted waiting for engineering. We can have a conceptual layout circulating within engineering in a few hours and out to the field by the next day. Field technicians can tag and disconnect terminals right away while we figure out how to wire up the new unit. Perhaps in a perfect situation you might get all this done in a few days with all hands on deck but that's more luck than aptitude.
mensetmanusman 7y ago
How long would NYC be out in the case of a massive solar flare that blows out all transformers in the city?
- WrtCdEvrydy 7y ago
  At 2-3 weeks a piece... let's just say we'll be eating twinkies for at least 2 years?
- ics 7y ago
  The interesting thing in this scenario is that if units were lost and unrecoverable in large swaths you wouldn't have to worry nearly as much about matching the legacy system perfectly, accidentally tripping out adjacent units, or things like that. Transformer replacements might even be easy, we'd just take whatever we had and wire it up anywhere it fits and can provide meaningful service. I'd be more worried about the issues to the underground system (nearly 100,000 miles of cable and piping) once oil stops circulating for long amounts of time, generators, and battery systems.
vageli 7y ago
> I work on one of the densest power grids in the world (NYC) on the team which does emergency transformer replacements and can possibly answer questions about this. From failure to in service for a single transformer in Manhattan is down to 2-3 weeks. That includes design and engineering (my part), permitting, and construction. Interesting. I had always wondered how utilities interact with local permitting. I had assumed they would work something out to have standardized permits for common work which they could leverage during incidents but that seems that's not the case.

by ics 7y ago

MisterTea 7y ago

Neat job. I've always wanted to poke my head into ConEd's system and see the whole thing from behind the scenes. Years back I remember ConEd was replacing two or three very large transformers down the street of a large venu I was working an event. Turns out we worked the same event the next year and amazing the same day ConEd again had the second transformer scheduled to moved into place. What an operation. Huge trailer from Bay Crane along with one of their giant hydraulic cranes. Was told the transformer weighed in at 240,000 pounds.

ics 7y ago
Heh, our most recent replacement a few weeks ago came in at almost 900,000lbs with oil. Depending on the speed of the replacement you can't always move the transformer dry because you need a certain amount of time between filling it and putting it in service.
- jacobush 7y ago
  Air bubbles need to rise first?
- MisterTea 7y ago
  That sounds like quite the transformer. That must have been at one of the big 345kV subs. And I assume shipping filled is to avoid the time needed to remove air bubbles. I know a few people in the ConEd trenches, one guy is a lineman who does all the >=69kV stuff and the other guy worked in substations and was in mott haven last I spoke to him (cushy job). I'm always amazed how much new tech con ed has installed and how much old stuff is still kicking around. I live in south queens and we're all 4kV overhead fed from area subs in turn fed by 27kV. Some of those 4kV poles have westinghouse transformers from the 30/40's still in service, glass insulators, and leftover wooden pins from older insulators. Some are so old the primaries have fallen off and just hang between two adjacent poles, some the crossarms are half falling off. Amazing this grid doesn't go down more. Then the next pole is brand new, composite strain insulators, cutouts, reclosers, metering, etc. All modern hardware. And I know some areas they upgraded the 4kV overhead to 27kV. The secondaries are in even worse condition, ancient cloth insulation half weatherd/baked off. But it all still works. I really love how there's this massive network of cabling and systems which orchestrate the flow of power from multiple synchronized generators so it can get to your humble wall socket and charge a phone or turn a light bulb on.
  - snowwindwaves 7y ago
    They are shipped without oil and full of nitrogen gas to reduce the weight.

ethbro 7y ago

Are transformers like these standardized, or are they generally built as one-off customs for the specific site? (I'm not an EE, so not sure how many variables one needs to contend with in grid-scale transformer design)

ics 7y ago
We have purchasing standards which are particular to our system, though specs may match others' by coincidence. Most of these are electrical, e.g. how the primary and secondary windings are configured (delta-wye in most cases) but also what gets mounted to the outside (bushing potential devices, current transformers, gas monitoring units, etc. These are not unique to each site which is why there can be so much work to replace one. Older stations were actually designed around the transformers in some cases and the physical configuration of a 50+ year old failed unit is rarely like those produced more recently. We might have a unit on hand that is electrically equivalent but impossible to install without essentially rebuilding the entire station. To give a rough overview of the range of transformers on our system, the first thing we care about is the voltage and MVA rating, then whether the bushings are open air or gas insulated, what the winding configuration of the primary and secondary are (delta-wye being most common), whether the unit it is replacing has a tertiary winding (sometimes at transmission voltage, other times distribution or for light and power to the station), and then maybe what accessories it comes with such as current transformers for protective relaying.

asdfasgasdgasdg 7y ago

I'm curious how long that process would take if it were do or die. Obviously, permitting would no longer be involved. So the real question is how long is the critical path through engineering->construction?

ics 7y ago
It varies significantly from project to project. What we depend on is the accuracy of existing records (30-60 year old microfilm scans typically, schematics and physical drawings), similarity of the failed and replacement units, ease of access to the site, support needed by other disciplines (e.g. concrete or steel work), and cost even if it is do-or-die. A perfect like in kind replacement could possibly be done entirely by construction using proper procedure of documenting their disconnects if there were no structural modifications required to move the units in/out. This is rarely the case however and so we need to check everything from phase clearance on making the high voltage connections to whether disconnecting a failed transformer will trip out another perfectly healthy one, possibly making the situation much worse. All that being said, we make every attempt to triage the work required and get information to construction as soon as possible so that no time is wasted waiting for engineering. We can have a conceptual layout circulating within engineering in a few hours and out to the field by the next day. Field technicians can tag and disconnect terminals right away while we figure out how to wire up the new unit. Perhaps in a perfect situation you might get all this done in a few days with all hands on deck but that's more luck than aptitude.

mensetmanusman 7y ago

How long would NYC be out in the case of a massive solar flare that blows out all transformers in the city?

WrtCdEvrydy 7y ago
At 2-3 weeks a piece... let's just say we'll be eating twinkies for at least 2 years?
ics 7y ago
The interesting thing in this scenario is that if units were lost and unrecoverable in large swaths you wouldn't have to worry nearly as much about matching the legacy system perfectly, accidentally tripping out adjacent units, or things like that. Transformer replacements might even be easy, we'd just take whatever we had and wire it up anywhere it fits and can provide meaningful service. I'd be more worried about the issues to the underground system (nearly 100,000 miles of cable and piping) once oil stops circulating for long amounts of time, generators, and battery systems.

vageli 7y ago

> I work on one of the densest power grids in the world (NYC) on the team which does emergency transformer replacements and can possibly answer questions about this. From failure to in service for a single transformer in Manhattan is down to 2-3 weeks. That includes design and engineering (my part), permitting, and construction. Interesting. I had always wondered how utilities interact with local permitting. I had assumed they would work something out to have standardized permits for common work which they could leverage during incidents but that seems that's not the case.