With all the posts about building a generator, it shook loose a question about transfer switches. I understand why the xfer switch is there, and the implications and safety hazards to lineman by not having one. That said, if the "Main" breaker is thrown open, why couldnt the generator be wired in to the panel through another breaker (of appropriate rating) that is 'downstream' of the main breaker? If this is unclear, basically imagine having 2 "main" breakers on the same service panel, one leading to the grid and the other to the generator. You are effectively isolating the entire system from the grid, and so long as the generator is shut off and breaker kicked open before the main breaker is rethrown to reconnect to the grid there wouldnt be any problems, correct? I realize that all this would have to be done manually, but this would be used in emergency situations only (no machines running, just lights, heat, fridge, etc). I would use grid power at all other times. I would also provide a mechanical link between the breaker throws to ensure that if the generator breaker is being closed, the grid main must be open to allow the other to close.
I am not opposed to getting a true transfer switch, I just dont see why this approach is not equally save and effective. Is there something I am forgetting?
"Forgetting" is exactly the problem with such a setup.
I am not opposed to getting a true transfer switch, I just dont see why this approach is not equally save and effective. Is there something I am forgetting?
While it might be reasonably safe WHEN OPERATED PROPERLY, it is not designed in such a manner that it always WILL BE.
You may not always be the one to operate it. You may not always be thinking 100% clearly. If you sell the building, the subsequent owners may not have any idea of the proper sequence of operation.
A proper transfer switch is as close to "foolproof" as you can get. It is simply NOT POSSIBLE to accidentally connect the generator to the utility lines with such a setup. So without one, you run the risk of either backfeeding the grid, endangering lineworkers, or having your generator connected to the line when utility power is restored, destroying the generator.
Not to mention that anything other than a proper installation is a NEC violation, and an excuse for your insurance company to cancel your policy in the event of ANY electrical or fire-related claim.
Even a mechanical link between two breakers or disconnects isn't foolproof, whereas the x-fer switch is as close to foolproof as you can get.
We installed a lot of centrifugal chillers when we were still in the contracting business, and they all use a star-delta motor starter. Via some contactors in the starter, it "rewires" the motor leads between the start and run modes. These contactors are both electrically and mechanically interlocked. Even with all that, we had one that was less than a week old where both the electrical and mechanical interlocks failed at the transition, effectively shorting all 3 legs of 480V together at the same time. This starter was fed from the MDP by a 400 amp breaker. If you looked at the starter, any rational person would swear it could never do this. I was standing about 5 feet from the starter when it happened, and once we made sure the power was killed and opened the starter door, it looked like a bomb had gone off inside the enclosure. Anything depending on mechanical or other interlocks can fail, but the way in which a manual transfer switch is constructed does not employ interlocks, so its as foolproof an electrical switch as you'd ever find.
The transfer switch provides a positive mechanical interlock to avoid closing in the generator breaker with the "main" breaker also closed in. The bigger danger in this is that a standby generator, backfeeding thru into the local distribution lines will also backfeed the transformers. Result is the "high side" of the transformers will be energizing the dirstribution lines with whatever distribution voltage is (could be 33,000 volts, 13,800 volts, or a few other common ones). For this reason, the National Electrical Code and any local electrical codes will call for a transfer switch with a positive mechanical linakge to insure there being no way that BOTH breakers can be closed at the same time.
In answer to your question: YES, a person can use a standby genset by FIRST opening the main breaker to cut any tie to the local distribution lines and then closing in the breaker connecting the genset to their service panel. Understand I am in no way recommending this method or offering it as an alternative to the use of transfer switches with positive mechanical linkages between the generator and main (line) breakers.
I have a setup I use in my own house for standby power. I do not endorse or recommend this method as it is not in accordance with NEC or any other electrical codes. It consists of a 60 amp fused safety switch with a 60 amp receptacle out in the garage. This switch is wired to a 220 volt breaker on my panel in the basement. Normally, I run the air compressor and a 3 HP Cincinnati Bickford drill off that 60 amp receptacle in the garage. I have a Lincoln gasoline driven welding machine with 6.5 Kw of 220 volt 1 phase power available. When we do have an outage that goes more than a couple of hours, I use it for standby power. Here is my method:
1. Get a flashlight and go in the basement and OPEN the main breaker.
2. Go into the garage & make sure the 60 amp safety switch is OPEN.
3. Roll out the Lincoln Welder. I have a heavy lead cord made up with appropriate 220 volt male plugs at each end. This lead cord is "mobile Home Cord" and good for at least 60 amps. The cord is kept ready to go, hung on a rack by the garage door. After the welding machine is rolled outside, I plug in the lead cord to the receptacle on the welding machine and to the receptacle in the garage. I then hook up a grounding jumper from the welding machine frame to a ground stake.
4. I start up the welding machine and get it running on "high idle". After it smoothens out, I plug in a 110 volt floodlight to get some base load on and get the generator governor to take over.
5. I then close in the 60 amp switch in the garage. This hits the welding machine with whatever loads in the house happen to be there- refrigerator, freezer, well pump, and oil burner all come on as a block load. The welder pulls it.
When we are hanging ont he welder, we do not run it continually. We run perhaps 30-40 minutes to an hour every few hours. This is enough to charge the well pressure tank for toilet flushing, and keep the food cold and the house warm. We cook on a propane range, so do not need electricity to cook with. We have an assortment of Coleman liquid-fueled lanterns (bought for 1-5 bucks apiece at yard sales) for lighting and these are kept fueled with good mantles, pumped up and ready to go on a rack in the garage. We use the Coleman lanterns for steady light. If the outage goes overnight, we keep a couple of the old Dietz wick type kerosene lanterns going with the wicks turned low for night lights. We also keep about a full cord of dry wood in the basement, so keep a fire in the wood stove thru the winter as well. There is no way that a Lincoln Wleder with an air cooled engine is a base-load generator.
The tricky part is knowing when the power has been restored. We lived away from any other houses, though a couple have been built in sight of us, so I suppose we will look for their lights. If you live in a remote spot as we did, the thing to do is to wire up a couple of indicator lamps on the incoming side of your main breaker. If those lamps come on, it means power has been restored.
One time, years ago, we had a windstorm outage and were out for about 12 hours. It was night time, and we had had our supper and the Lincoln Welder was powering the house. I had fueled it not too long before. Suddenly, the lights started flickering and going bright and dim. I could hear the engine on the welder starting to "hunt", which was uncharacteristic as it simply conks out when the fuel runs out. I ran out to the garage and opened the 60 amp switch taking the welder off line. I then went into the basement and close din the main breaker and found the power had come back on. In thinking things over, I realized the neutral leg of the wiring between my house panel and the distribution grid is hard wired- the main breakers do not disconnect it. I thought I might have had a backfeed from the grid via the neutral leg when the power was restored. I asked an electrical engineer about it and he told me that when the power was restored, the slug of load on the system caused a phase-to-phase imbalance on the primary (high side, 3 phase power) side of the distribution transformers in my localized area. This had the effect of putting a momentary current into the neutral leg and backfeeding the welding machine. Weirdly, single pahse transfer switches do not usually break the neutral leg, as this is a rather uncommon phenomena.
To give an example of a transfer switch gone wrong: I was on a job in Venezuela years ago. A telex (predated faxes and no cell phones back then) arrived at the jobsite, asking me to travel to Trinidad to handle an emergency job involvinv a melted-down generator. It was a bit of a job to get from the rural jobsite int he back reaches of Venezuela to Trinidad, but I did. I arrived at the Nelson Street Telephone Exchange a day or so later and was shown about a 350 Kw diesel genset. The engine looked fine, but the generator was burnt up and had molten copper coming out some of the vent openings on the stator. A new generator had been shipped from England and my job was to change out the damaged generator and make sure that genset was OK. The story of what had happened was pretty simple: a transfer switch the people at the Nelson Street Telphone Exchange had malfunctioned following a power outage. The gist of it was this: The electrical engineers there had designed their own transfer switch. What they used was two independent circuit breakers with automatic clopsing/opening by means of heavy springs. The springs were kept "charged" by small electric gear motors following each breaker operation. The breakers were worked remotely by control circuits. The phone exchange engineers had designed and built circuitry which sensed a 10% or greater voltage drop on the incoming grid power (or a loss of it) and opened the main tie breaker to the grid. At the same time, the diesel set was auto started. When it had built voltage, the circuitry then closed in the generator breaker. No mechanical interlocks between grid tie breaker and generator breaker. When grid power was restored, the circuitry sensed it and a time delay relay kept the place ont he generator for perhaps 20 minutes more to make sure the power was really back. After 20 minutes, the generator breaker would be opened and the grid tie breaker closed in. The generator was then run unloaded for another 20-30 minutes to cool it down as it was a turbocharged diesel set. This system had worked fine for years until a malfunction occurred. AN outage had occurred and the gneerator had come up and on line, carrying the telephone exchange. When grid power was restored, the time delay realy ran out and called for closing in the grid tie breaker. Unfortunately, some component in the control circuitry failed to provide the input that the generator was still on line with the generator breaker still closed in. The result was that the grid backfed the generator. The odds of the generator being synchronized (phased in) to the grid were so remote as to be impossible. The grid, having megawatts of power behind it simply overpowered the generator. The diesel engine was turnign the generator and the grid was trying to twist it into phase. The diesel had enough beef not to let this happen, so the windings melted down. I installed the new generator, checking the diesel for mechanical damage in the process. The diesel was fine. I advised the electrical engineers at that phone exchange to go get a transfer switch with a positive mechanical linkage between the grid tie breaker and the generator breaker to prevent a recurrence. I took a flight out of there the next morning so never heard any more about it.
There is a reason for using a transfer switch with a mechanical interlock, and it is to make things as relaible and absolutely foolproof as possible. I am not one to talk, using the system I do. I didn't hav eone installed when we were building our house- too costly and things happening too quickly. Now to install one would mean having the meter pulled to de-energize the incoming feed. The system I use works fine as long as it is done "by the numbers", without deviation.
Interesting story, Joe. I wonder what that
generator looked like as the molten copper was
oozing from the vents on the unit? Must have
been a sight.
I wonder if somebody really wanted to be double
careful when performing your trick, there is
one more thing they could do.
Once the main breaker is switched off, then
all of the load breakers could be opened as well
at that point. Then one of the electrical
testers (wigglers) could be used in continuity
mode to see that both busses on the panelboard
read infinite ohms to ground. That way one
could be dead-certain that the main breakers
had opened up for real.
At that point the main could be locked out
with a tab on the panel, and the load breakers
re-connected and back-energized from the garage
as per usual.
Pics of the switch that I installed. With it I can't forget to disengage the main breaker because I don't have to.
Next pic is where the generator plugs in outside. From here the wire runs to the switch panel in the pic above and plugs into the receptical on the front of the panel. Then I simple throw the six switches and the main elec. panel is isolated from generator power.
Have neede it several times in the last couple of years.
I believe the power industry term is visible break. The way I interconnect my 1.5 kw generator provides that.
I only use the generator to power the oil burner, the refrigerator, and several lamps. That is all it will power. This is probably not practical for more than light residential, but it serves my needs.
I replaced the oil burner switch at the burner with a outlet connnected to the house breaker panel and short cord. Normally I run the oil burner off that outlet.
During a power outage, I set up the generator outside, and run an extension cord through a piece of pipe running through the outside wall. I string the extension cord down the hall in the cellar till I get to the oil burner, where I put a 3 way connector ("cube tap") and a short cord. I unplug the oil burner from the power line, and plug it into the generator. The oil burner now gets power from the generator, and is isolated from the power line. I run another extension from the cube tap up the cellar stairs. I unplug the refrigerator and a couple of floor lamps, and plug them into the extension cord served by the generator. In all cases, the generator and the devices supplied by it are isolated from the power line. When the power comes on, the lights in the other part of the house come on, and we know to put everything back the way it was.
This could get tedious for a large scale installation, or where there are a lot of outages. There one needs a transfer switch.
FWIW, I have been told by an electrical engineer coworker with utility experience that in some cases at least, utility crews ground a line prior to working on it, due to capacitive coupling from other lines. (This was in the context of somebody forgetting to remove the ground before they closed the breaker back in.) If the line is grounded, your generator will not backfeed for very long.
Finally, I have had experience with one popular brand of residential circuit breaker. I have found that residential grade breakers do not take kindly to being used as switches. Sometimes they do not close back in.
I've installed and operated several generator/changeover switch installations in my career, all three phase and heavy amps.
I have to agree, that it is a bad idea to use breakers, no matter how they are connected. Even if the connection linkage is perfect, breakers often fail and they can fail in the closed position.
Every changeover switche I have ever seen consisted of a large knife switch that had two sets of contacts ON OPPOSITE SIDES of the enclosure. By having only ONE set of arms (knives) to make contact with the two sources, they are almost foolproof. Almost!
I definitely do NOT recommend it, but if I were going to make a home brew changeove, it would definitely follow that kind of design. Of course, a large knife switch, a motor, limit switches, gears, box to put it in, etc. would cost as much or more than a commercial changeover. All in all, it's not worth it. Just buy a real one and be legal and safe.
You could wire this similar to a reversing motor starter.
Grid power ON uses the Grid Contactor and blocks the coil of the Genset Contactor (thru a NC auxiliary contact on the Grid Contactor).
So, when you experience Grid power OFF the NC contact will allow the Genset Contactor to energize and supply power to the circuit panel.
As soon as the grid power ON returns, the Grid Contactor will re-energize and block out the Genset Contactor again.
The "start & run" signal for the Genset can also come off the same NC auxiliary contact.
This would take some large NEMA starters and a wall full of electrial panels. I certainly love the idea but the mechanical interlock is so much simpler and just as safe IMHO.
If a mechanical linkage interlock isn't enough, the double redundancy will have to have the contacts electrically "monitored" to pick out a failure of the mechanical linkage.
That kind of spec-grade work doesn't come free but it certainly is reliable and robust.
I work at a powerplant, so am a little familair with line crews. As was noted, whenever a line crew gets set to work on a powerline (unless it is "hotline" work), they do a few things without fail:
The first is to make sure the line is isolated. This is done by pulling disconnect links or fuses up on the poles using a "hotstick".
A "hotsick" is a fiberglass insulated pole with a variety of tools which can be mounted on it to work on conductors, buss work, transmission line hardware, etc which may be energized ("hot"). Hotsticks are tested with high voltage regularly to be sure the insulating value is not compromised.
The second is to sweep the line to be worked on with a voltage detector, also mounted on a "hotstick".
The third is to hang grounds. Grounds
are basically pieces of what amounts to welding lead which have special clamps to grab the conductors or hardware (such as taps on a transformer), and may then be shorted together (phase to phase) and run to ground.
The fourth thing is to call into the system operations (control room of nearest powerplant, or load dispatch or independent system operator, etc) and formally log the grounds on.
These steps are followed without fail to insude the lines to be worked upon are dead, wtih no possibility of stray currents or induced currents finding their way into them. Unfortunately, if someone accidentally backfeeds a line with a standby genset, they will likely "saturate" or charge the windings of any transformer between their genset and the lines being worked upon. The effect is only momentary as the genset will be run against a shorted/grounded line on the high side of the transformer. While the lines are grounded out, the removal of the ground can produce an effect similar to working a set of ignition points on an old style ignition system. There may be a flashover of transmission or distribution voltage. Linemen do things by the numbers, so should be using hotsticks, face shields and insulated gloves to take off the grounds. Regardless, an incident of this type, with a flashover is serious, and there will be an investigation. The person who backfed the grid with a standby genset is going to find themselves liable for any injury or damage resulting.
FWIW: Specially designed circuit breakers are used around powerplants and switchyards and substations to switch higher voltage electrical loads on lines and heavier electrical equipment. Breakers used in lower voltage service ( under 600 volts) are generally molded case breakers which, as noted, really are not designed to be switched under load. The arcing at the contacts when the circuit is madeup or broken under load will damage the breaker if it is done as routine thing. Low voltage Molded case breakers do not have provisions for repeated operations under load.
For occasional switching under load, an enclosed safety switch (basically a ganged set of open knife switches) is an economical alternative.
Manual transfer switches are offered for sale by a firm called "Allen Manufacturing". They advertise in "Boats and Harbors". Another firm which may have manual transfer switches is "Groban" out of Chicago- an industrial/military surplus sales firm. These r are rotary transfer switches and are good for most home/light commercial standby gensets. I think they are rated at about 100 amps. Who/where these switches are made & whether these switches are UL Listed is something I don;t know.
In addition to all the interlock wiring, a proper automatic transfer system will also incorporate time delay relays to make sure that the genset is up and running before cutting over to it, and to make sure that grid power is stable before dropping the generator offline. Most commercial units also incorporate timers for automatic cycling of the generator for testing, and allowing the generator to run for a short time period with no load for "cooldown" after a cycle.
Add in voltage, phase rotation, and frequency monitors, and you end up with quite a lot of control equipment in addition to 2 big contactors.
Just a question, supposing that a genset did backfeed the grid, wouldn't the load out there immediately pull the voltage down to a point where the prime mover couldn't maintain voltage or cyclic rate?
I don't know how the controls are setup on a genset, would the engine just die or does it chug along producing 69 volts at 32 Hz?
In answer to your question: if a standby genset were to backfeed the grid, assuming:
-No grounds hung
-No load on the "high" side of the transformer
-Disconnect links or high voltage fuses pulled to isolate the portion of the distribution grid where the damage or outage had occurred
Then: the genset would be pumping voltage thru the service connection and thru the first tranfromer (whether on a pole or on a pad or in a vault). This would be putting full distribution voltage thru the grid.
If there were loads on the grid int he form of other houses with their main breakers and individual circuit breaker closed in, then the standby set would be hit with all that load.
The result would depend on the degree of protective circuitry on the set. Assuming a typical, simple standby genset without any such under voltage/underfrequency protection, the genset would lug down horribly and voltage would collapse. I do not know if this is the "69 volts 32 Hz" you are referring to. Assuming the set were backfeeding through a 200 amp main breaker on a household service panel. If this were the case, the set would lug down and backfeed the local grid, but only for the transient time it took the breaker to heat up and trip.
Most portable standby gensets are designed for isolated buss/single unit operation. They are not designed to be paralleled with the grid or even another genset. Heavier gensets may be built with protective relaying to allow parallel operation of multiple gensets. Getting a genset to parallel with another required being able to "phaes in" or superimpose the sine waves of the AC current each genset is making so they line up before the breaker on the genset being put on line is closed in. That is putting it simply. Aside from that, the genset being paralleled or put on line needs to have some means of regulating the load the engine carries to prevent reverse current motorizing the generator. When a governor is holding an engine's speed to maintain frequency on a generator set operating as single, isolated unit, it acts as a governor continuously. This is called "isochronous" governing. When the genset is paralleled with others or into a grid (known as an "infinite buss"), the greater power of the bigger gensets or grid takes over and maintains the frequency (frequency determines rpm). The governor on the smaller genset then serves to control load. The governor resumes "governing" when the breaker tying the genset in with a buss with other generators or the grid is opened.
A funny story comes to mind. It was not funny for the guy who lived thru it. It happened about 28 years ago. I was working for a dealer/OEM in heavier medium speed diesels and steam turbines. This dealer/OEM also sold high speed gensets- driven by 'Cat, Cummins, Detroit, or gas turbines. They built up custom gensets and built a line of brushless AC generators in their own plant. This firm was selling a lot of gensts and equipment to the Saudis. I refused to go to any Arabic nation, and was running in and out of South America with the bigger medium speed engines and an occasional steam turbine. One of the other guys got sent over to Saudi Arabia with four packaged gensets. These units had good sized 'Cat diesels and brushless AC generators made in the dealer's own plant. The package had been designed as a kind of portable power plant for isolated bus operation. The design was such that the four units could parallel with each other and power up an isolated and finite grid. The four gensets and the control panel and relaying had all undergone full load testing on a load bank at the dealer's plant.
The sales contract called for a startup engineer to go over to Saudi and commission the plant and train the local operators, so another guy got sent over there. I saw him a few months later when I got back from a trip to South America and he told me what went on. It seemed the gensets had been trucked out to some town and set up. My friend the startup engineer had done the wire checking for all four sets and the control panels. The sets were started and "phase checked" (for correct phase rotation so all the motors int he town didn't run backwards). Final connections were made to the buss into the local grid. For a few days, things ran well. Then, the Saudis started piling additional loads onto the grid. My friend said this was mostly A/C loads. This was block motor starting loads and really puled down the grid voltage. My friend kept telling them they had run out of capacity but they would not listen. The four gensets could not hold up voltage.
A Saudi electrical engineer arrived and told my friend that he would solve the situation. My friend said the Saudi engineer had a superior attitude, and there was no reasoning with him, so he let him go at it. The Saudi engineer figured if the voltage was collapsing, paralleling the four 'Cat gensets with some more grid power would solve the situation. My friend told the Saudi engineer that there was no way that the four 'Cat powered gensets were going to be able to be paralleled into an infinite buss/grid as they didn't have any sort of synchronizing equipment for synching with the grid, nor was there sufficient protective relaying. The Saudi engineer pretty much belittled my friend and took over. He had a contractor quickly run a pole line from some prong of an existing grid, bringing the pole line into the location where the four gensets were. The Saudis rigged up what they thought was appropraite instrumentation transformers and a synchroscope, but located it several hundred yards from the gensets. The big day arrived and the Saudi engineer and his operators went to parallel the heavier grid with the four 'Cat powered gensets. My friend said the Saudi crew was running around yelling in Arabic, with one guy at the gensets to work the breakers manually. There were Arabic operators lined up to yell up and back as to whether things were in phase, and whether to close the breaker in. My friend said it was bedlam and the Saudi operators were not remotely knowledgable about what synchronizing a generator to a buss was all about. At that time, any gensets going to Saudi used to be ordered with markers on the gauges and meters to show what correct readings or correct ranges should be.
With much yelling and screaming and arm waving, they did close in the tie breaker. My friend said all hell broke loose at the four 'Cat powered gensets. Smoke and fire flew and the engines all shut down automatically. Four generators had been destroyed as had the control panel just that quick. It was a combination of reverse current, and being horribly out of phase (unsynchronized).
My friend went to his hotel and telexed the dealer/OEM in the States. He got set to go back to the USA as he figured there was nothing else he could do. He never made it to the hotel lobby. In came the Saudi engineer with a couple of Saudi police officers and placed my friend under house arrest. The Saudi engineer told my friend the whole mess with the four diesel gensets was his fault and the fault of the US Dealer/OEM. My friend denied this and was told he would be held under arrest until the US dealer/OEM made good on four new gensets, a control panel and some other equipment. My friend kept telling the Saudis the problem was due to their disregarding the fact the gensets had not been built for "infinite buss" or main grid power operation. After a few days, my friend was taken to a Saudi police station and the Saudi engineer told him to write out a "confession" admitting the gensets to be improperly designed, made with substandard materials and workmanship, etc. My friend refused. The Saudis really threatened him and finally escorted him back to his room with a warning to "confess". About 48 hours later, he had another "confession" shoved accross the table at him. He was told if he didn't sign, he would not see home again. He signed it and was put on the next flight out.
The Saudis took the confession and arrive din the USA with a US law firm and raised hell including pulling some diplomatic strings. They insisted that even the 'Cat diesel engiens were shoddily built and wanted four entirely new gensets and another control panel plus monetary damages. With their diplomatic connections, they got waht they demanded. FOur new gensets were built and these got enough protective relaying to put them in the class of a base load utility power plant in the USA. My friend did NOT go back over there for that startup. I think they stuck some poor startup engineer from the UK with the job.
Interesting story Joe. Memo to self: 'never,
ever, ever, get sent to Saudi Arabia.' Thank
you for making it abundantly clear why.
What bad incident instilled that principle in
you, before the disaster that happened to your