Ask me about industrial electricity 2: electric boogaloo

The Something Awful Forums > Discussion > Ask/Tell > Ask me about industrial electricity 2: electric boogaloo

grover: Jan 23, 2002; PEW PEW PEW

zapradon posted:

Oh sorry, thinking about too many things at once. It was a 9390, transformerless.

I'm told it's fairly similar to the 9395.

Yeah, the 9390 and 9395 use virtually identical technology. I think they even share some of the same controller cards. And you do need to be very careful to handle the neutral properly; they have a dedicated set of IGBTs to derive the neutral voltage, and it can be thrown all out of whack of you don't provide a solidly grounded wye source.

Edit: for those of you trying to follow along, this is what we're talking about.

grover fucked around with this message at 02:25 on Feb 22, 2012

# ¿ Feb 22, 2012 02:19

Adbot: ADBOT LOVES YOU

# ¿ May 21, 2024 05:53

grover: Jan 23, 2002; PEW PEW PEW

Three-Phase posted:

You have to be REALLY CAREFUL if this thing is already online and powering mission-critical loads, but can you bypass the UPS? I believe with some Powerware systems, there is an option that allows you to bypass the UPS so that you can do maintenance safely while your critical load is still energized. Just to make sure the problem is directly related to the UPS and not your power line.

You'll see these circuit breakers marked "MBP and MIS". Be sure to read Eaton's instructions extremely carefully if you are running a critical load like a production server, an iron lung, nuclear reactor control rod servo motors, etceteras so you don't get in trouble/kill people.

Grover, isn't there also a way in software to bypass the UPS, or is that to just shut it down? (The last time I shutdown an Eaton Powerware 93XX at work, despite running all the required paperwork, resulted in some "unexpected fun".)

It's not only possible, but standard practice to install redundant UPS systems so that you can secure one while operating on the other(s), and to have a maintenance bypass switch so that you can bypass the entire UPS system for major maintenance or replacement or whatever reason. (Most companies like to transfer from utility power to emergency generators while on maintenance bypass so that an errant utility failure doesn't take their facility down.) All high-end double conversion UPS systems also have what's called "static bypass", where a solid-state static switch (often in conjunction with a mechanical contactor) can engage in the event of an internal component failure and this ensure continuous power to the critical load, even if the UPS has failed. To transfer to maint bypass, you always place the UPS in static bypass first. Otherwise, the UPS would end up with the output feeding the input once the maint bypass breaker is closed, and possibly out-of-phase since the UPS output isn't necessarily paralleling the bypass input source (though it usually is). As maintenance is required on all model UPSs, transfers like this are routine and done safely all the time.

Transformerless UPS units present a new challenge in implementing this. Most notably, the lack of a transformer and single-model-economies-of-scale means lack of a 208V model 9395; they're all 480V. (As are comparably sized Liebert NX/NXL and APC Symmetra PX UPS.) Put a 208Δ-480Y step-up transformer on the input, and a 480Δ-208Y step-down on the output, and now you've fixed the voltages and have your neutral, but your maint bypass is 60� out-of-phase with your UPS's output and you get a nice big BANG when you try to transfer. In the old days, it was easy: you turn the first transformer backwards* and go 208Y-480Δ then 480Δ-208Y. +30� and -30� cancel, and you're in-phase. But you can't do this when you need a solidly grounded neutral. Therefore, you need to use custom-wound Δ-zigzagY transformers with 0� phase-shift to keep your UPS in-phase with maintenance bypass switch.

*To do it right, you'd get it custom-wound, as you always want the transformer primary coil inside the secondary for highest energy efficiency and lower magnetization current, though an off-the-shelf delta-wye wired-up backwards would work in a pinch.

grover fucked around with this message at 04:17 on Feb 22, 2012

# ¿ Feb 22, 2012 04:03

grover: Jan 23, 2002; PEW PEW PEW

zapradon posted:

I've been agitating for a Fluke 435 without much success so far..

We've got a couple Fluke 435s and a Dranetz for power monitoring. They're bulky enough that I don't generally travel with it it unless I think I'm going to need it, which means I never have one when I do

So, I've started keeping a Fluke 345 in my tool bag now. Not quite as good as a 435, but still gives incredibly good info and I'm much more likely to actually have it handy. I keep flinging it around like it's a $20 multimeter, though, and am terrified I'm going to forget myself and smash it...

grover fucked around with this message at 00:04 on Feb 23, 2012

# ¿ Feb 23, 2012 00:01

grover: Jan 23, 2002; PEW PEW PEW

ncumbered_by_idgits posted:

I am curious what kind of attitude facilities of a similar size have in regard to electrical safety.

The attitude I've seen is one of widespread and chronic complacency. They'll wear PPE if you call them them out on it, but get pretty bitchy about it.

After all, the shock-hazard of a 2000A 208V panel is no different than a 120V receptacle. :suicide:

# ¿ Mar 25, 2012 01:46

grover: Jan 23, 2002; PEW PEW PEW

ncumbered_by_idgits posted:

I wasn't specific enough. I meant what is management's approach?

Full oversight in theory, not so much in practice. Lot more blissfull ignorance and cover-your-rear end poo poo than actual enforcement. I try to do my part to set things right when I can- last thing we need is someone getting hurt.

# ¿ Mar 25, 2012 02:19

grover: Jan 23, 2002; PEW PEW PEW

some texas redneck posted:

There's lots of pad transformers around campus, generally labelled with the building name that transformer serves and what I assume to be the amperage - ranging from 20A to 200A depending on the building.

The union, for example, has a 100 amp service, if the bigass transformer is to be believed (it's about 4-5 ft tall). It's a massive 4 story building with a hydraulic elevator, has its own chiller, etc. Kinda curious what kind of voltage would be going in and out of those transformers, I'm assuming there's another transformer in the building to knock it down to 480.

Transformers like that are typically labled in "kVA". 100kVA ~ 277A at 208/120V 3-phase. The transformer you're describing sounds more the size of a 500-750kVA transformer, though. Can you read the nameplate on it?

Three-Phase posted:

To be honest, my biggest fear in this industry is not being shocked, killed, or ending up in the hospital with the poo poo burned out of me in an oxygen tent. My biggest fear someday is lying in my bed at night thinking "I told John it was OK to work on that system live. His wife and little kids will never see him again."

One of our electricians was electrocuted a few years ago. At home, working on his computer. Really tragic.

Left a wife and two kids behind. Really brings home the dangers of complacency, though.

grover fucked around with this message at 15:30 on Mar 25, 2012

# ¿ Mar 25, 2012 15:14

grover: Jan 23, 2002; PEW PEW PEW

DaveSauce posted:

Since the subject is on arc flash, what's protocol for separating 480v power and 120v control in the same panel in order to lower PPE requirements?

It's good practice to separate controls from 480V conductors to increase safety when maintaining the (often maintenance intensive) control components. Arc flash is literally an explosion. Picture a fireball engulfing the 480V enclosure, exiting through every slot, every hole, and blowing out panels and any weak spots. The more power available, the bigger the explosion. You'd need to have an engineer run the calculations on available fault current to determine how big the "boom" will be, and what someone in front of the 120V portion would be exposed to, with mitigations in place. Your goal seems to be to reduce it below what PPE they're already required to wear.

Two boxes nippled together should work pretty well, but what you're describing would probably only really suffice for reducing the shock hazard, not the arc-flash hazard.

grover fucked around with this message at 20:53 on Mar 25, 2012

# ¿ Mar 25, 2012 20:51

grover: Jan 23, 2002; PEW PEW PEW

Three-Phase posted:

Grover - are there any things like Lexan shields that could be installed in a cabinet to mitigate arc flash? (I doubt it, I've never seen anything like that, except for "touch-proof" shielding.)

Yes, those can work well in many cases. I see them a lot more in Japan than the US for some reason, but when used properly, it's fantastic. (That said, I've not actually seen the calcs so maybe it's not actually rated... )

Arc-flash explosions look like simple explosions in video online, but in actuality, it's comprised of plasma which is heavily impacted by electromagnetic fields, and doesn't necessarily work like you think. I've heard stories of metal doors blown clean off a switchboard and thrown into the opposite wall, yet an insulator just inches away come through without so much as discoloration.

# ¿ Mar 25, 2012 23:21

grover: Jan 23, 2002; PEW PEW PEW

Not an Anthem posted:

I was just browsing this thread because I'm going to ask some electricity questions when I get my bandsaw revamped, but I just bought an old 40's or 50's bandsaw that features that exact SquareD switch and my friend was joking that it will somehow kill me. Forgive the utter ignorance but is this switch a bad thing or a better thing than a previous bad thing?

That SquareD switch is fine, so long as it's in good condition. It was an exposed knife switch that killed poor Jones.

# ¿ Mar 26, 2012 16:39

grover: Jan 23, 2002; PEW PEW PEW

Three-Phase posted:

And for one last treat - that sweet sound of transformers starting.

I love how that guy puts on all that arc-flash gear yet doesn't bother to properly velcro the front of his suit, leaving it hang open.

# ¿ Apr 7, 2012 00:32

grover: Jan 23, 2002; PEW PEW PEW

Three-Phase posted:

As a fun little aside - I've heard of some factories where ties are strictly prohibited, at least on the factory floor (must be removed or safely "tucked inside" a person's shirt). The reason being that with moving and rotating machinery, a tie can catch and fling/strangle/pull in a victim.

Legend has it that clip-on-ties were invented for safety.

# ¿ Apr 9, 2012 22:07

grover: Jan 23, 2002; PEW PEW PEW

Three-Phase posted:

I see (and make) lots of prints in C and D size.

We don't have any laser plotters. I do most of my prints in A size :smithicide:

We do have an inkjet plotter cable of full-size color prints, but it's sooo slow. I've actually come to prefer A-size to half-size because it's way more convenient to store & carry. There are some drawings just need to be really huge to be intelligible, though. Especially when the draftsman ignored all standards and used the smallest font possible.

# ¿ Apr 10, 2012 02:21

grover: Jan 23, 2002; PEW PEW PEW

Three-Phase posted:

Unfortunately for legal reasons I'm not going to make statements about the products.

This is my personal view - any extraordinary claims require extraordinary evidence or explanation.

It's a complete and utter bullshit scam, and I would testify in court to that.

# ¿ Apr 18, 2012 00:11

grover: Jan 23, 2002; PEW PEW PEW

Three-Phase posted:

Yeah, I've found the emergency reporting leaves something to be desired in a lot of situations.

That generator starting/stopping is a problem. You would think that the emergency system would be configured to run for a few minutes after it detects power has been restored so that it doesn't keep stopping and restarting the generators. The disconnect switches I've seen have a timer where the power must be restored for like at least 15 seconds before it switches over then shuts down the generators.

Timers are also typically set for 3-5 seconds before the generator starts/transfers to ensure that power is actually "out" and not just a brief blink. They often do have other relays that will start the generator and transfer if a certain # of brief outages occur within a certain time frame. Another timer is usually set to continue on generator for 15-30 minutes after power returns to ensure that it's stable. They also run in idle a few minutes after that to cool down.

If they were actually starting up/shutting in rapid succession like that, they were set horribly wrong. In this size range diesel gen, they can start and come up to RPM and voltage in about a second and shut down very quickly as well. They shouldn't be doing it repeatedly, though!

grover fucked around with this message at 19:47 on Apr 20, 2012

# ¿ Apr 20, 2012 17:12

grover: Jan 23, 2002; PEW PEW PEW

A good % of power engineering jobs are inherently local in nature, and very unlikely to ever be outsourced. I'm not so sure of the advantages of an MS, though. A BS plus 2 years work experience would probably serve you better than an MS.

# ¿ Apr 20, 2012 20:03

grover: Jan 23, 2002; PEW PEW PEW

Three-Phase posted:

53GVA.

Assuming it clears in 3 cycles... if I'm converting right... it's roughly equal to a 1000lb bomb.

# ¿ Apr 21, 2012 16:30

grover: Jan 23, 2002; PEW PEW PEW

Ender.uNF posted:

How is that even possible without everything exploding?

SF₆ (Sulfur Hexafluoride) insulates 2.3x better than air at 1ATM, and is even better at higher pressures. It also has excellent arc-quenching characteristics. Put simply, sparks and arcs don't travel well in SF₆, so you can put high voltage conductors much closer to each other than you otherwise would be able to. It has a net effect of reducing the volume of high voltage components by a factor of about 10.

It's also non-toxic, doesn't burn, and conducts heat well. It's all-around great for high voltage components.

grover fucked around with this message at 14:05 on Apr 22, 2012

# ¿ Apr 22, 2012 14:01

grover: Jan 23, 2002; PEW PEW PEW

Three-Phase posted:

Didn't see this until recently:

ABB Develops 1.2MV Circuit Breaker

1200kV Dead Tank Breaker

Wow. The end connector has five separate corona rings.

I don't think I've ever seen a diagram of a high voltage interrupter. What do they look like inside?

5 corona rings is a bit insane!

# ¿ May 12, 2012 12:26

grover: Jan 23, 2002; PEW PEW PEW

Cichlidae posted:

So, what kind of safety equipment are we talking?

hahahaha, no, you don't have to wear that for a typical 120V cabinet. You'd have to have an engineer run calcs for available fault current to be sure, but you're probably looking at class 0 (non-melting street clothes like jeans and a long-sleeve cotton shirt, safety glasses, and leather gloves). At worst, you'd need to don 4 or 8cal flash-rated clothing that's otherwise indistinguishable from a pair of dickies along with a face-shield hardhat and leather shoes.

For the actual regs, check out NFPA-70E. The rules vary depending on what you're doing and the arc fault potential. You need one level to open the door, another if you're standing 3' back when someone else opens it, something else if you're connecting equipment, and the worst-case only if you're actively grabbing and working with energized conductors.

grover fucked around with this message at 01:04 on May 16, 2012

# ¿ May 16, 2012 00:57

grover: Jan 23, 2002; PEW PEW PEW

the nicker posted:

So does anyone know about arc fault (AFCI) breakers?

I just bought a new house and it came with arc fault breakers, which so far I've had to get 3 of them replaced with regular breakers because they kept tripping during normal operation.

My electrician claims that they kind of suck. He says he puts them in there to start with because it's what inspectors like to see, but usually he'll end up having to come back out and replace some of them because they trip when they're not supposed to. Is he full of poo poo?

Your new house is probably full of wiring errors, no doubt due to your electrician miswiring poo poo. They should NOT be tripping during normal use, and it's illegal to remove required AFCIs and replace them with normal breakers. Tell him to fix the loving wiring faults or you'll report him to the city and have his license revoked.

If he suspects a bad AFCI, swap it out with another. Test with known good power sources, like simple lamps. If it trips when you plug a lamp it, it's most likely a ground-neutral fault.

grover fucked around with this message at 22:00 on May 16, 2012

# ¿ May 16, 2012 21:58

grover: Jan 23, 2002; PEW PEW PEW

the nicker posted:

Is there anything I can check myself to verify this before calling them back and accusing them of being idiots? I did of course have the house inspected and the electrical passed without issue. He tested the ground at every outlet and also poked around inside the breaker box.

The first one would trip randomly when I had my entertainment system going (tv, receiver, cable box, ps3 etc).

The second one would trip if I flipped on the flood lights at night (2 sets of motion-sensing floods with 2 150-watt bulbs each).

The third was in the kitchen and would flip if I plugged the vacuum into one of the outlets (vacuum worked fine in any other area of the house).

What specifically could cause that? Other than a faulty AFCI, but seems weird that 3 out of 8 or so would be bad right out of the box.

edit: Come to think of it, the three AFCI breakers that were tripping were all for circuits that also had GFCI outlets. The kitchen, guest bath (same circuit that my entertainment wall is on), and master bath (i remember them telling me that the flood lights in the back were also on the GFCI circuit in the master bath). Coincidence? I'll verify that when I get home. However in each case it was the AFCI in the breaker box that was tripping, I've never had a GFCI outlet trip.

Sounds to me like the electrician done hosed up somewhere. Whether the GFCI or the AFCI would trip would depend where the wiring fault is.

You could swap a known good AFCI with one that's tripping and see if the "good" one starts tripping and the "bad" one suddenly works fine. That's always a good indication of an issue in the circuit. You could also disconnect the ground wire from the panel and see if there is still continuity with the neutral- that's always a good sign of a "he done hosed up".

# ¿ May 17, 2012 22:41

grover: Jan 23, 2002; PEW PEW PEW

Three-Phase posted:

BONUS! MV SWITCHGEAR FROM MERLIN GERIN!
Photo from Schneider Electric

This is some of the weirdest looking gear I think I've ever seen.

MGE sells components to 3rd parties who fabricate the switchgear in their garage (often figuratively, surprisingly often, literally, though). There seems to be little rhyme nor reason nor consistency between design of MGE switchgear in various parts of the world; it's at the whim of whomever is fabricating it. It makes you really appreciate how well we have it with standardized mass-produced gear here in the US.

One of Schneider Electric's regional headquarters (of a multi-national overseas region I'll not name) is literally in a house, as in a housing development mcmansion in the middle of a residential suburb, with normal homes on all sides. Apparently, zoning laws are lax in that country and it was cheaper than an office building. Even has a garage to do the work in right on-site!

grover fucked around with this message at 02:36 on Jun 7, 2012

# ¿ Jun 7, 2012 02:33

grover: Jan 23, 2002; PEW PEW PEW

We had a transformer explosion at a sister site recently. The dielectric in the oil-filled transformer had degraded from a leaky cover plate and arced over when it was re-energized following routine PM. The explosion threw the electrician literally 15' through the air, landing in a parking lot about 25'. Fortunately, he was wearing full PPE and walked away with only minor injuries. Wear your PPE, kids! And don't skimp on fluids testing, either.

Jonny 290 posted:

For a bit of content, a question: Are there enough advances in the technology that there's a rift between old-guard guys that want coal-powered steam driven switchgear and new school kids that want to put everything on computers or whatever? I imagine at this point it's kind of sink or swim as far as adopting new tech, for the old fellas.

Not really much choice: all the new stuff comes with computers built right in. It's far more reliable than the old stuff, and has less maintenance, and I've yet to hear of even the oldest dog pining for mechanical relays or poo poo line that.

It's driven a lot of specialization on the engineering and technician side of the house for design/maint/repair, especially for controls, but there really isn't a whole lot of difference in cabling or conduit or lugs or any of the really time consuming stuff that occupies electricians most of the day. The new solid-state equipment operates different internally, but all hooks up pretty much the same way.

grover fucked around with this message at 17:44 on Jun 8, 2012

# ¿ Jun 8, 2012 17:39

grover: Jan 23, 2002; PEW PEW PEW

I got to tour Surry Nuclear Power Plant today. Was really loving awesome climbing around all through the turbine building around the massive piping, heat exchangers, generators, etc., for the two 800MW steam turbine generator systems. Didn't get to go into the reactor containment, building, though

grover fucked around with this message at 01:32 on Jun 23, 2012

# ¿ Jun 23, 2012 01:28

grover: Jan 23, 2002; PEW PEW PEW

Allantois posted:

Grover,

Did you see the Generator Step up transformer for that?
I used to work for a power plant design firm but we never got much over 500MVA Transformer.

Curious as to if the standard on that size of unit would be for 3 single phase transformers or to really have a single 3 phase transformer rated at like 1000 MVA?.

Also interested in how big the isolated phase bus duct is between the steam turbine generator and the GSU. I'm guessing those generators cant be generating much more than 25kV, which would put the current at around 23kA, assuming these are 0.8 pf machines!

That is one hell of a conductor. IPBD has always amazed me.

Unfortunately, I didn't. The generators (800MW) each appeared roughly the size of a 40' containers, and run at 22.5kV. Voltage is stepped up in the to 230kV and 500kV transmission voltages, but we never made it over to the switchyard.

The reactors are in the breast-shaped buildings; the turbine building is the large building directly south of them. You can see the switchyard pretty clearly, just to the southeast of the turbine building:

Google maps link

I found it interesting that the water the plant returns to the James River is only 2 degrees warmer than what it takes in. Also, the intake canal is apparently awesome for fishing because little fish get sucked into the pumps but then grow too big to fit through the gratings and because they have no predators, just get MASSIVE. And the output is a favorite fishing spot, too, because fish thrive so well in the warm water there.

# ¿ Jun 23, 2012 02:31

grover: Jan 23, 2002; PEW PEW PEW

Allantois posted:

Based on that picture it looks like three single phase transformers which is not too surprising. If you notice just to the south of the turbine building the three transformers for each of the turbines. These transformers have to be directly adjacent to the turbine because of the MASSIVE current delivered between the generators and the GSU's. That IPBD is probably around 55-60 inches in diameter per phase. You can see it if you look closely.

I asked the engineer giving us the tour about it, and I thought he said the transformers were in the switchyard, but I might have misheard. The lines you can see in google maps were guyed from the building about 2/3 of the way up, but there were upconductors from near ground level. I could never get a good view from where they actually came from.

The generators themselves were fully enclosed, and on the top floor of the turbine building alongside the turbine assemblies (6 turbines per generator- 2 high pressure and 4 low pressure). With the maze of piping and heat exchangers inside the building, I wasn't able to trace any of the HV cabling within the building. It's amazing just how many ancillary heat exchangers are hanging everywhere to wring an extra few % thermodynamic efficiency from the steam.

Since power is so critical to nuclear plants, even when they're shut down, there is probably an auxilliary set of transformers designed to backfeed the building from the grid when the turbines and EDGs are offline.

grover fucked around with this message at 03:29 on Jun 23, 2012

# ¿ Jun 23, 2012 03:21

grover: Jan 23, 2002; PEW PEW PEW

Three-Phase posted:

How many poles were the generators? Eight or more? Did you take a look at the excitation system? Was the cooling air, water, or hydrogen?

The generators themselves were completely enclosed giant cylinders that looked (to me) to be roughly the size of a 40' container. So, no clue on the poles or excitation. I couldn't see what they were filled with, but this plant was designed in the late 60s (commissioned 72 & 73), so I doubt it's very exotic. The turbines were entirely enclosed as well.

Allantois posted:

Regardless that is a really cool plant. I wish I could tour it.

Plants that are lucky enough to have a cooling water source like that and can dump it right back into the bay save a ton of $ on cooling systems. I notice there are no hyperbolic cooling towers so iconic to what people think of when they think of a nuclear power plant.

It's no coincidence; they site power plants like this specifically for ease of cooling. A large % of the turbine building is dedicated to the massive heat exchangers cooling the secondary loop.

Edit: Dominion's website has an animation that looks like a pretty accurate depiction of the Surry power plant, as far as # of turbines goes, layout of the components, etc:
http://www.dom.com/about/stations/nuclear/nuctour.html

grover fucked around with this message at 03:52 on Jun 23, 2012

# ¿ Jun 23, 2012 03:34

grover: Jan 23, 2002; PEW PEW PEW

Three-Phase posted:

I don't think the two are there for the purpose of redundancy, because even if you power off one transformer via the isolation switch, you'll backfeed the connection via the other transformer. It would be very important to NEVER do any maintainance downstream of, say, the LEFT open switch with the RIGHT open switch closed. (Grover - let me know if this assumption is correct.)

I'm kinda stumped, too. My first thought was that it was a soft-starter (as wye connections are often used as soft-starters), but that doesn't explain the cross-connect. Then I was thinking that maybe they used both delta and wye windings in each motor to change them from 3-phase to ~ 6-phase like you see in a lot of 12-pulse power supplies (only other time I've seen transformers like that), but that still doesn't explain the cross-connect.

Notice, though: no means of isolation on the secondary. These transformers are ALWAYS in parallel. I think they used two transformers due to some limitation that prevented use of a single larger transformer.

Qs. Are the motors redundant, or sequenced that only one will ever be at (or near) full load at any given time? Are they linked together mechanically, in which case they could act as zero-point starters for each other?

grover fucked around with this message at 10:44 on Jun 27, 2012

# ¿ Jun 27, 2012 10:42

grover: Jan 23, 2002; PEW PEW PEW

Three-Phase posted:

The thing is that it's possible to prevent the grid from completely becoming overwhelmed by doing things like rolling blackouts and brownouts. What's bad is if something unexpected happens when the system is already very stressed.

Is that why 75% of Puerto Rico doesn't have power right now? I mean, it's kinda occam's razor, but the news doesn't have much by way of details right now.

# ¿ Jun 30, 2012 00:24

grover: Jan 23, 2002; PEW PEW PEW

Commercial requirements have lots of variables, but DoD standards are that all rooms, including data centers, comm equipment rooms, power rooms, etc., must have wet pipe sprinklers. (Dry pipe preaction sprinklers are only permitted in very special conditions, like unconditioned mechanical rooms where the pipes could freeze and burst.) The theory is that the sprinklers are only going to activate when there's actually a fire, and even then, only the sprinklers nearest the fire are going to activate- it's not like hollywood where the whole building goes up; each sprinkler head is only going to activate when the glass cylinder holding the valve shut gets hot enough to explode, and when that happens, you've probably pretty much write off anything under it already anyhow.

Modern dry-gaseous fire protection systems (like FM200) are often used in sensitive areas with lower thresholds for activation to extinguish fires before the sprinklers discharge. They're not legal as the sole protection system, though; sprinklers are still required. But they're awesome for limiting water damage. It doesn't smother the fire like CO2, but chemically interrupts the flame. If you're in the room when it goes off, it's not going to kill you, but it's apparently extremely uncomfortable, to say the least. The video above is great and all, but you're not going to be reading a book through it.

The easiest solution if you're worried about water damage to switchgear is to spec everything with a NEMA-3 enclosure, but that kinda defeats the point of having sprinklers. In all but the most sensitive applications, any electrical system subject to a deluge sprinkler should be interconnected into the fire suppression system to shut off during a discharge anyhow. When it's all said and done, it's a lot easier to replace damaged equipment than damaged equipment AND a building that burned down around it.

grover fucked around with this message at 22:28 on Jul 13, 2012

# ¿ Jul 13, 2012 20:14

grover: Jan 23, 2002; PEW PEW PEW

Fire suppression systems (sensors, alarms, fire pumps, etc.) in a building are generally powered from a dedicated breaker upstream of the main breaker, so that even when the main breaker is opened, there is still power to the fire suppression system. It gets a little more complicated when you add generators and UPS systems in, but there are pretty strict NEC requirements for what you can and can't put on a legally required emergency generator system. Sometimes, multiple generator systems are required so that one can be shut down while the emergency system stays up. It gets even more complicated when you have a "continuous industrial process" which is exempt from the shut-down requirements, but still requires protection if the fault/fire is somewhere within the continuous industrial process which can't safely be shut down.

The interconnection between the fire suppression system and other safety systems like emergency power off buttons to provide an emergency shut-down signal is generally just a set of alarm contacts. The criteria for a shutdown is not necessarily when the alarms go off (fire alarm systems are notoriously whiney and always having nuisance alarms), but when something more fundamental occurs, like the flow sensor in a sprinkler pipe feeding a particular room sensing a discharge.

grover fucked around with this message at 16:15 on Jul 14, 2012

# ¿ Jul 14, 2012 16:11

grover: Jan 23, 2002; PEW PEW PEW

some texas redneck posted:

So a neighborhood in the next city had a transformer fail. From the way the news is describing it, it sounds as if the transformer windings may have shorted.

This is one of the relevant articles. Thankfully it was limited to 6 houses, though there was 1 person injured from a light catching on fire.

So how can dry soil cause a transformer to fail in such a spectacular fashion? And would a (good) surge protector have done any good in saving anything? And finally, assuming the houses saw 1kV or above, what else would be damaged? Breakers, wiring, etc? I'm assuming they saw at least 400-500 volts, since compressors got fried. The houses are new enough to have AFCI breakers, I would assume those got toasted along with GFCIs.

edit: VVV looks like a good movie

A bad ground can cause issues, but this doesn't sound like a bad ground. It sounds like it was a short in the primary windings that quite simply put out excessive voltages on the residential side.

A high impedance ground could potentially have caused a lot of heat, but the symptoms would be dimming lights, not melting lights. I think this excuse is a cop-out.

grover fucked around with this message at 16:35 on Jul 28, 2012

# ¿ Jul 28, 2012 16:23

grover: Jan 23, 2002; PEW PEW PEW

some_admin posted:

Theives have stolen all the ground wires on our utility poles, for a couple of blocks around.
They cut them 1' and 6' above ground.
What is the risk to my household electronics ( and mylife, I guess)?

I'd unplug everything sensitive during a thunderstorm (physically unplug, don't just turn off) until they get it fixed; otherwise, you don't need to worry too much about it.

You may also want to avoid touching the grounds that haven't been stolen, as they're likely carrying a bit more current than usual and may be at a higher voltage above ground than normal.

grover fucked around with this message at 19:17 on Jul 28, 2012

# ¿ Jul 28, 2012 19:13

grover: Jan 23, 2002; PEW PEW PEW

The only reason old facilities don't have it is because it wasn't invented yet when they were built; it's common to the point of ubiquity in newer facilities. MOVs are cheap, and there's no reason not to have one on every panel. Standard practice is to install one SPD at the main panel/service entry to protect from lightning strikes, and to install additional SPD at subpanels to protect against internal surges as well as provide additional protection against lightning strikes.

Motors and switching transients are simply horrible for generating damaging transients in machine shops and other industrial buildings. The tools themselves can often take it, but the electronic controls can't.

FYI, you can get a whole-house residential surge suppressor for about $20 for your own home, and probably should.

grover fucked around with this message at 13:47 on Jul 29, 2012

# ¿ Jul 29, 2012 13:36

grover: Jan 23, 2002; PEW PEW PEW

Three-Phase posted:

Voltage up, current up: issue with the source (surge)
Voltage down, current down: issue with the source (sag)
Voltage up, current down: source stiffness issue at the source or issue with the load
Voltage down, current up: issue with the load (overloading or motor starting) or source stiffness issue at the source

I've seen those last two listed as being load-only problems. But I'm not sure I agree with that. (Grover - what is your take on my assessment here?)

Resistive loads will see voltage and current go up/down together. Motor loads will see current decrease when line voltage increases because less current is needed for the same kW/hp output at higher voltage. Non-linear loads depend on how the power supply is designed, but can be tracked by the harmonics. You'll see voltage drop at increased currents regardless of the type of load, with motor starting current and transformer magnetization current common causes of voltage sags/dips.

Unless it's something stupidly obvious, like the only large load in a small facility, tracing the specific source of a power quality issue can be very difficult. Especially intermittent ones.

# ¿ Jul 29, 2012 23:06

grover: Jan 23, 2002; PEW PEW PEW

Arc Flash as a "big deal" has been a fairly recent phenomena in the US; most of it has been on the books for a long time, but it was never really prioritized. Europe will come around to it eventually, I think.

I don't like the way grounds are dealt with in much of Europe, either. It just seems wrong routing grounding conductors via a different path than the current-carrying conductors.

Three-Phase posted:

Where I work, with NFPA 70E, in a lot of cases you need Level-2 PPE to open the cover to even a 120/208V panel. A lot of people are chagrined by that. However on big electrical systems, you can get pretty fearsome arc flash levels even on low voltage panels.

US department of defense procedures requires Level 1 arc flash PPE to so much reset a 20A breaker on a 120V panel. It's ridiculous! For those not familiar, Level 1 PPE is special fire resistant coveralls over natural fiber clothes, face shield, leather gloves, hard hat and electrical resistant steel-toed shoes. Commercial standards for the same is safety glasses, leather gloves and non-melting/natural fiber clothes right on up until you physically start sticking a screwdriver to a live circuits inside a 120V panel.

grover fucked around with this message at 01:43 on Aug 5, 2012

# ¿ Aug 5, 2012 01:34

grover: Jan 23, 2002; PEW PEW PEW

FrozenVent posted:

Wait, isn't 20A at 120v what my house runs on?

Cause I've reset those mofos barefoot a few times.

That's what most of us in the real world do, too, yes. It's no wonder people don't take PPE seriously when the rules are so unrealistic about it.

# ¿ Aug 5, 2012 02:14

grover: Jan 23, 2002; PEW PEW PEW

Panaflex posted:

I have a 2kw AC to DC rectifier for the purpose of powering a Xenon projector lamp. The input requirements are 3 phase 240v AC 60hz at about 18amps . I only have residential 200 amp 240v split phase. I heard from someone once that it was possible to create a capacitor bank to generate a 3rd phase and drive the rectifier. Rotary converters are out since there is apparently too much ripple. Could you provide any insight as to how to assemble such a circuit?

You can't do it with caps, not with any real success. There are ways to do it, but they're more complicated than just using a cap, and it would be cheaper to sell the 3-phase rectifier and buy a 1-phase.

If it's just a "dumb" diode rectifier, you may be able to just hook up 2 phases and run in a derated (1.2kW) condition. No promises on the ripple current, though; it won't be as clean as if the rectifier had balanced inputs since it's designed to smooth a 360Hz ripple and you'll actually have 240Hz. Would help if you added new capacitors in parallel with those already in the rectifier, though.

E: series is not parallel

grover fucked around with this message at 12:27 on Mar 4, 2013

# ¿ Aug 7, 2012 22:52

grover: Jan 23, 2002; PEW PEW PEW

some texas redneck posted:

So how long are large capacitor banks like that supposed to last (even land based)? The document suggests 20 years, I see some stuff that's easily twice my age every time I drive anywhere though.

It depends on the capacitor, really. The electrolytic power capacitors I'm most familiar with (appx the size of 40oz beer cans) are recommended to be replaced every 5 years, but frequently last much longer. Failure of small non-replaceable electrolytic caps embedded into circuit boards is one of the leading causes of old-age death of computers and consumer electronics. Solid state caps and those with solid separators last just about forever. The Queen Mary 2 used Czech aluminum/polypropylene/paper capacitors immersed in dielectric oil, and should have had a very long lifespan- the manufacturer says 20 years. Dielectric oil requires maintenance, though, and the oil in these capacitors had moisture content well above typical values, reducing it's ability to insulate. Seven capacitors had been replaced on the QM2 since 2006- one was defective, one was arcing, and five suffered "mechanical damage."

Actual life of capacitors depends on the electrolyte and other conditions- heat decreases their operational life of some designs by quite a bit (many electrolytic caps have their operational life cut it in half for every 15F over 77F, just like VRLA batteries). They'll generally give warning signs before failure, in the form of reduced capacitance and swelling, but you have to be looking for it because it generally isn't readily apparent without specific instrumentation. They apparently were looking, as the seven other caps had been replaced pre-emptively, but in this case, the protective circuit for the faulty capacitor had failed with the display showing an unlikely (but seemingly ideal) 000mA current imbalance and nobody thought to look any deeper.

What most likely happened was swelling of a polypropylene insulator caused a contact to break loose, which caused electricity to spark across the gap, locally vaporizing some of the dielectric oil. Vaporizing the oil deep in the tank in the absence of air would not cause it to ignite, but while oil is a very good insulator and lets you place high voltage conductors closer than you otherwise would be able to, vaporized oil is not- get the bubble in a bad spot and it would arc between terminals like a full-on short circuit. At that point, BOOM!

QM2 Report posted:

Deterioration of the capacitors in the HF was monitored by measuring the current imbalance between the capacitors in each of the three phases.

...

It was not possible to establish whether the aft HF circuit breaker tripped due to an overload in either rank, or due to an imbalance in its rank 4 section as the AIS did not differentiate between the two types of alarms and recorded it as an �overload/ unbalance�. It is known, however, that within 300ms to 500ms of the alarm, the circuit breaker opened. Almost simultaneously or just before the circuit breaker opened, there were several �IO-FAULT� alarms (indication of loss of reactive power) at both high voltage switchboards and generators. There were also several �DISCREPANCY� alarms (indication that the switchboard had detected a loss of one phase) at several bus bars in the 690V circuit. No undervoltage alarms in either the high or low voltage circuits were registered by the IAS.

I've seen harmonic filters fail before and turn into charred masses, but this one had a lot of power behind it!

You can see the capacitor broke it's mountings and blew several feed upwards when it exploded:

QM2 Report posted:

The probable conditions that existed just before the accident: oil weeping or spraying out from the failing capacitor casing; 11kV voltage across the three phases; and the subsequent release of flammable gases from the capacitors, were ideal for an arc-flash event to take place. The melted corners of the copper bus bars indicate that arcing did take place between the phases, and it is therefore possible that the explosive noise heard by the crew members was due to the shock waves produced by the rapidly moving hot air emanating from the arc column.

It's very fortunate nobody was in the room when it blew. As to the cause, the report is full of finger-pointing. The cruise line is blaming the manufacturer for shoddy capacitors, but the manufacturer is blaming the cruise line for damaging them with a lot of transient surges. Or, as the report more diplomatically put it, "The fact that the unit did fail catastrophically indicates that either there was a system side problem overstressing the capacitor or that the capacitive elements could not perform to their stated rating."

grover fucked around with this message at 13:40 on Sep 4, 2012

# ¿ Sep 4, 2012 12:14

Adbot: ADBOT LOVES YOU

# ¿ May 21, 2024 05:53

grover: Jan 23, 2002; PEW PEW PEW

Two Finger posted:

Anyway, one particular point to note, is this:

During drydock, for whatever reason, the protections were not tested, so it was not established that they were nonfunctional.

Yeah, that's pretty damning. I'd be willing to bet it was actually discovered- or at least suspected- to have failed, but the tech doing the PM didn't want to deal with it.

Also damning is that one of the other capacitors failed a few years prior with an imbalance of 1250ma, an imbalance level well in excess of the 400ma alarm point and 800ma automatic trip point, and mere 50ma shy of the 1300ma it's calculated to arc over and explode! You'd *think* someone would have taken action to check & fix the CT and relay circuit at that point, yet nobody did. During the investigation, this CT was found to have an open primary coil and had been completely non-functional for literally years before the explosion. So, no protective circuits tripped until the exploding capacitor bank tore the electrical bus bar off the other capacitors in the bank and the subsequent arc flash triggered a different protective relay.

grover fucked around with this message at 21:14 on Sep 4, 2012

# ¿ Sep 4, 2012 21:07

The Something Awful Forums > Discussion > Ask/Tell > Ask me about industrial electricity 2: electric boogaloo

«‹›4 »