PDA

View Full Version : Making an automatic transfer switch



alternety
09-24-2008, 03:16 PM
In my ongoing project (called a house) I want to add a backup generator. I prewired for this. One subpanel has all the backed up circuits. I am going to used an inverter based Honda generator. It is not designed to be used as an automatic backup generator.

I have available a controller that understands the Honda and sees internal faults (e.g., oil pressure) and does the proper starting and stopping protocols. The problem is a transfer switch. I find only two types. Manual and automatic. The only automatic units I have found want to be the processor that runs the generator. But they don not know the internal details of the chosen generator.

So, what I need is an automatic transfer switch that will take commangd from the generator controller. The controller will sense power loss, wait to be sure it is not just a short term transient problem, start the generator, and when it is up to speed, transfer the suppanel. On grid restore it must transfer power back to the grid and cool down the generator.

OK. Got that. I need a transfer device. My thinking is a 3PDY contactor. A contactor is designed to minimize arcing and handle a bunch of current. I have spent a bunch google hours trying to find a device. There are more annoying terms used for these contractors than I am familiar with. The sites that sell them also tend to assume that if you don't know the manufacturer and part number; you really don't need to buy one from them. My guess is that what I am looking for is a reversing contactor (to get the 3PDT) but I am not sure. The current rating must be for the types of loads expected in a residential subpanel. Not resistive. To minimize cost I have been trying to find a surplus contactor.

Can anyone give me a hand in identifying what device(s) I should be looking for. Manufacturer and part number would be most appreciated. A surplus source. even more. I need 240VAC, 60 or 100A. There seems to be a real price premium to get 100A.

An additional question; what are the regulatory constraints of doing this?

jwelectric
09-24-2008, 03:35 PM
An additional question; what are the regulatory constraints of doing this?

It is called third party testing.
UL is one of the best know third party testers.
What you are proposing is nothing short of suicide

Speedy Petey
09-24-2008, 03:36 PM
I have two comments. Please don't be offended as this is simply my own opinion.

1) I think you are crazy for trying to "build" you own automatic transfer switch.
You say you found something that knows or senses what is going on inside the generator? I highly doubt that.
Transfer switches are made to work with specific standby generators. The switch is the brains and the generator is the brawn. They work together and only together.

2) I see you said inverter type. I was under the impression that the largest unit was 3000 watts. I just checked it out and they do make a 5k and 6k unit.
This makes a difference. I was going to say 3k is not nearly enough IMO to run a transfer panel.

alternety
09-24-2008, 03:59 PM
JWelectric. - A double pole relay is what is inside an automatic transfer switch. It is unable to be in two positions at once. I did see one unit that used two relays. It was capable of making a mistake. If what you said had any credence it would be murder, not suicide.

Speedy Petey - 1) Generally. But I would not have said what I did unless it was true. Why would I lie? And no; I do not simply not understand.

2)Yes.

Anyone want to actually help?

jimbo
09-24-2008, 04:11 PM
In their own gentle way, I belive the folks are trying to tell you not to do this....in other words a "home made" device is probably a code violation, in that it is not listed by UL for the use. I am sure you are well aware that the sensitivity on the transfer switch issue is related to the safety of linemen and the public. I believe people have been killed by home generators backfeeding a "dead" main. You don't really want to be involved in that.

By the way, a UL listing for your "invention"......I think runs about $25,000.

Speedy Petey
09-24-2008, 04:18 PM
I'd be very curious to see this controller then. I have never seen nor heard of a generic controller available to the general public.
Or maybe you have an inside track......?

alternety
09-24-2008, 04:38 PM
It is not generic. www.atkinsonelectronics.com

Actually it probably is pretty generic. All you have to do is map the engine signals for oil temp, etc. to the inputs and deal with their start circuit. But they have already done the generator of interest to me.

Bill Arden
09-24-2008, 04:40 PM
There is one way to do what you want within code...
1. Install a large outlet.
2. Install a generator panel that lets you switch circuits over to a generator plug. This type of panel has wires that go from each panel breaker and wires that connect to each load that used to go to the breaker.

The switch is then not part of the house wiring and thus not subject to UL requirements when used by the house owner in his/her own house .

I used this method to feed UPS power to a light circuit.

-
I'm not sure what controller you are talking about.
Personally, I like the Atmel AVR and ARM processors. ;)
The coldfire based NetBurner boards are nice too.

-
A simpler option is to use a standard automatic panel and trick it into waiting for the generator to be ready and to switch on when you want it to.

alternety
09-24-2008, 05:01 PM
Bill, yes you see the issues. The manual switch is however not automatic; which I need to keep food frozen if I am not home.

Yes the control of an ATS would do it. But it would significantly increase the cost. That is why I did not mention that in this thread. It is a last resort in terms of cost. I would need to know a bit about how any specific ATS works to try to fool it. Some of the generator vendors are real difficult about explaining how their stuff works. There are some generic ATS's out there. It would probably work to intercept their start signal and start the generator with the other controller. UL is a significant advantage of this approach.

Bill Arden
09-24-2008, 05:22 PM
Bill, yes you see the issues. The manual switch is however not automatic; which I need to keep food frozen if I am not home.

Yes, but what I'm saying is that the manual generator panel would be permanently in "generator" mode and you would provide power either from the inverter or the big outlet.

It's also interesting to note that you might be better off buying a UPS since they already have a "transfer switch" and an inverter that can run off battery's.

alternety
09-24-2008, 10:17 PM
I am not sure you understand "inverter" in this context. The inverter I am talking about is internal to the generator. The generator behaves just like any normal generator in terms of output. The merit is very clean power and a variable speed engine tied to actual power consumption. Forget that inverter.

There are no batteries with an associated inverter.

Here is what I would really really like. A set of grid attached inverters (say OutBack) with internal transfer switches which produce the generator start signal this controller prefers. One side of the ATS to grid and the other to the inverter. With a battery bank this gives me millisecond transfer (effectively a part house UPS). With this scenario I would have to rectify the generator output in a battery charger (which also needs to be smart). If I void the warranty big time, I might be able to get some sort of DC from the generator. I would need another transfer switch if I wanted to do anything else. Now this inverter/battery configuration would be slick. But very very expensive. Have you priced big deep discharge batteries lately. One online store mentioned they had gone up 300% in, I think, 5 years. And still going. They are also sort of a pain in the butt. Assuming flooded cell (which still seems the best $/AH deal) you need an enclosure, venting for H2 and acid fumes. They have to be watered and the charge/discharge closely controlled. A modern charge controller handles the charge part. The inverter does discharge (probably).

Even slicker is to add some renewable energy source. I have no hydro potential. I have insufficient wind resources. I have real questionable solar prospects. I live near Seattle. We don't actually get sunlight. The state flower is a moss covered mushroom. And, hills and trees significantly shorten my sun day on the rare occasions there is some. Some compensation could be had by using a very big solar array. Can you say $$$?

So, with everything considered, I am trying to grid connect with a generator backup not designed to do this.. And we are back to the beginning of the thread. The generator is expensive (about $3500) but is quiet and has the best operation fuel consumption of any standby generator I have found. This is important for operational cost, but more importantly, in case of a long term power outage. A bit low on capacity, but my numbers say it will be OK. Some loads may have to be managed, but they are not connected at this time. There are going to be a couple of 5 HP motors eventually, but I can do without them in an emergency. For long outages the battery bank would be useful in that the inverter could handle the house and when the batteries got to 20% (or max of 50%) discharge, the generator would kick in at max capacity, charge the batteries, and go back to sleep. This may or may not minimize fuel consumption. But it should run the generator for less hours. The losses due to conversion in the charger, conversion in the inverter and cycle losses in the chemical processes in the batteries may cancel the fuel savings of full capacity operation for a shorter time. Depth of discharge is a MAJOR determinate in battery life. You really don't want to go below 50% and less is much better.

Another issue I have not mentioned with many, if not most, ATS boxes is the inclusion of distribution panels and breakers. This is already in place. All that it does is add cost. And it may prevent a single 240V circuit from using the full capacity of the ATS.

I have thought this out and analyzed a whole bunch of stuff. My question is about a relay. Period. Access to the relay can be creative.

jimbo
09-25-2008, 06:40 AM
"" The generator is expensive (about $3500) ""

You mentioned a relay rating of 240 volts/100 amps. That's 24KW. You think $3500 is expensive for a 20+ KW generator??

Most of us here live in the real world, not the back-woods survivalist world. You are welcome to your project, but notice that we just can't offer much in the way of the type of help you need. It's not in our genes!

Bill Arden
09-25-2008, 09:54 AM
I have priced deep cycle battery's and had an opportunity to buy a 10Kw UPS and some clear tank battery's, but the power utility would not let me use the off-peek meter to power the unit. The unit would have been able to handle the 4 hours per day that the off-peek meter turns off, but apparently regulations have not caught up.

I still think your best bet is to separate the house wiring from the project.

Here is what I would suggest.
1. Separate non backed up loads from backed up ones to reduce the amps needed. A separate Sub-Panel would work good for this. The high amp loads would be in the main panel. (Electric stove, Electric Hot water heater, welder)

2. Add a generator breaker bracket to the sub-panel to "side power" the panel. This switch would be manually set to generator when you aren't working on it.

3. Add a 50 amp 4 wire outlet

At this point your automatic transfer switch would plug into the 50 amp outlet and would provide power to the side breaker of the sub-panel.

This allows you to manually switch the sub-panel back to the main for maintenance, or upgrades, or if you sell the property.

I could take photos of my setup if you are interested.

PS: I also wired my generator to apply the same 120 volts to both phases so that all 240 volt loads don't draw any current.

Bill Arden
09-25-2008, 10:13 AM
Most of us here live in the real world, not the back-woods survivalist world.

I can't think of any reliable survivalist power sources in Seattle.
1. cloud Solar is too expensive.
2. Wind is an option, but not very continuous.
3. I'm not sure about ground source heat power since your outdoor temperatures don't change as much as they do here in MN.

4. I don't see any high temp geothermal sources near you.
5. It does not sound like you are in a area where you can access the ocean for tidal or a thermal difference power.

That just leaves bio-fuels, wood, natural-gas, and oil for Seattle.

alternety
09-25-2008, 11:36 AM
The generator is 5KW. The sub-panel is wired for 100A from the main panel. It would be easier to leave that alone. But I may reduce the breaker there. The loads on the sub-panel should be fine at 30A. The sub-panel is specifically configured to be only the backup loads.

My original intent was to get a generator in the 15KW range, but that seems unnecessary. There is also a very large jump in price between 60A and 100A breakers/ATS. I will run wire for 60A to the outside to allow for any future upgrade.

Bill Arden
09-25-2008, 12:00 PM
The generator is 5KW. The sub-panel is wired for 100A from the main panel. It would be easier to leave that alone. But I may reduce the breaker there. The loads on the sub-panel should be fine at 30A. The sub-panel is specifically configured to be only the backup loads.

My original intent was to get a generator in the 15KW range, but that seems unnecessary. There is also a very large jump in price between 60A and 100A breakers/ATS. I will run wire for 60A to the outside to allow for any future upgrade.

It looks like all you need is a manual transfer switch bracket for that sub panel and to add a 50 amp breaker from the main panel

It's ok to power the sub panel using a 50 amp outlet, 60 amp wire and a 50 amp side breaker. The key here is that the outlet has a 50 amp breaker and that all the house wiring is UL certified.

On the other hand, any transfer switch that is installed between the sub panels "main breaker" and the main panel has to be UL certified and rated for 100 amps.

That is why I'm saying it would be cheaper and easier to route the power to a smaller transfer switch threw a smaller breaker and then run the power back in the side of the sub panel.

Note: The small bracket on the sub-panel prevents the main breaker from being turned on while the side breaker is on.

Edit: The other advantage of this side-breaker plan is that you don't have to pull an electrical permit every time you change something since you can use 50 amp outlets to divide the "house wiring" from the transfer switch and generator.

Note: You will have to run a separate hard-wired ground wire between everything, but that does not require changing the house wiring.

Here is an image of a metal bracket Interlock for a sub-panel.
http://www.interlockkit.com/images/MEPanel033a.jpg

maintenanceguy
09-25-2008, 04:01 PM
If you install a UL listed generator with it's UL listed ATS according to all the rules and codes and something goes wrong and it kills somebody, you don't go to jail.

If you self-engineer the most brilliant system ever seen by man without following all the rules and codes and without UL listing and it kills someone, you go to jail.

For the same amount of work and less liability, you could invent a freezer that will stay cold for longer in a power outage. I see CO2 tank with a solenoid valve that opens to trickle liquid CO2 into a dry ice generator when the power drops.

alternety
09-25-2008, 04:44 PM
Do you perhaps have a link to the UL procedure for testing ATSs? Or the stamped metal plates that prevent you from manually turning the breakers to the wrong position before the other has been changed (kind of like a DPDT relay)?

jwelectric
09-25-2008, 07:06 PM
Do you perhaps have a link to the UL procedure for testing ATSs? Or the stamped metal plates that prevent you from manually turning the breakers to the wrong position before the other has been changed (kind of like a DPDT relay)?

This is contained in the UL Standards and one must buy a copy of the Standard in question. Access is not available without cost. See UL Standard 1008 for transfer switches.

jimbo
09-25-2008, 07:58 PM
First step in the procedure: take the item to be tested to a UL Certified Testing Facility. second step...pay the fee....Approx. $25,000

Bill Arden
09-25-2008, 11:09 PM
For the same amount of work and less liability, you could invent a freezer that will stay cold for longer in a power outage. I see CO2 tank with a solenoid valve that opens to trickle liquid CO2 into a dry ice generator when the power drops.

*grins and says sarcastically* (no offense intended)

1. If you use R-744 Freon (Co2) or vent it without a license you will get a big fine.

2. You need a license to re-fill that Co2 tank.

3. Storing pressurized Co2 Tanks in a residence is also dangerous.

*sighs*
One thought would be a brine and water ice tank. There is a site (I can't remember the link) where this guy shows how his heat pipe based freezer works. It freezes ice during the winter and then stays frozen all summer.

Scaling this down could be as simple as packing some of those "blue ice" packs in the freezer and bottles of something in the fridge.

What I'm saying is that people have to chose between buying off the shelf items that are UL approved OR they can do "research" and build what they want to as long as it's not part of the house wiring.

I've done UL testing and it's more about liability and not about product safety. I once had a UL approved power strip short out and smoke. It was poorly designed and could have done more than smoke, but it was UL listed since the chances of it starting on fire was low.

We sometimes have to ask ourselves where we draw the line between safety and cost. In Europe for example they decided that outlets must be safer than here in the us. There outlets are huge and expensive. Here in the US on the other hand we allow people to drive while talking on cell phones...

alternety
09-25-2008, 11:47 PM
Oddly I have had the same experience with a surge strip. Manufactured by a major computer company, UL rated. Power did not go out; the lights dimmed. The strip blew smoke all over the place. It did not start a fire (hurray UL) but I am not sure that it would not have done so with some paper dropped on it.

The reason I asked about UL procedure for the product is because I was curious if they did anything that actually evaluated the safety function of the device or just that it would not catch fire.

I don't disagree about the liability aspect of no UL label.

jwelectric
09-26-2008, 03:33 AM
I would think that the job of a Nationally Recognized Testing Laboratory would go a little future than checking to see if something caught on fire.
One of the labs for UL is located in Research Triangle Park in Raleigh, North Carolina just a little over an hour from where I live. I get to talk with some of the persons that work with UL on a regular basis, see the Education Committee (http://www.nciaei.org/section/committees/) and look for my name as well as Mark Ode. Once you see that we sit this committee together then Google search Mark Ode or just click here (http://www.ul.com/regulators/ode.cfm) Mark is one hell of a nice guy and will bend over backwards to help anyone he can.

The one thing that no NRTL can do is idiot proof anything that they test. A power strip will be tested at it rated amperage and voltage. Should someone buy one and load it in excess of its rating it will surly be damaged. Should someone buy a power strip and abuse it by throwing it in the floor and letting anything and every thing fall into it surly it will fail. Any of these would not be the fault of the NTRL but the fault of the end user.

I am sure that the power strips was never tested as they are being used below;

http://lh6.ggpht.com/_OKhT0exLkIY/R1i3X6D6QbI/AAAAAAAAAFg/DYyEOU0V1uE/dscn0198.jpg

maintenanceguy
09-26-2008, 03:44 AM
*grins and says sarcastically* (no offense intended)

1. If you use R-744 Freon (Co2) or vent it without a license you will get a big fine.

2. You need a license to re-fill that Co2 tank.

3. Storing pressurized Co2 Tanks in a residence is also dangerous.

Ok, we're way off topic here but I use one of these on a fairly regular basis. I have an EPA refrigerant cert but have never needed it to get a tank of CO2.

But I'm not serious about using one to keep a freezer cold in a power outage. No more than I would recommend designing your own ATS and connecting it to the power grid.:rolleyes:

Redwood
09-26-2008, 05:27 AM
OOOPS!
All the savings of making your own just went Bub-Bye!

Mikey
10-06-2008, 04:59 AM
If you do ever build your controller, here's another feature to consider...

I have a 16kW generator and 200A whole-house transfer switch setup all UL-approved and installed by licensed electricians. When utility power fails, the controller waits a while before starting the generator, then waits for the generator to warm up before transferring the house to generator power. Fine. However, when utility power returns, WHAM! -- the house is immediately transferred back to utility power. And the airconditioner controller hard-stops, requiring a service procedure to reset it. Nobody knowlegeable from Carrier has ever told me why this happens, but the installer, AC repair guy, and I kind of agree that the transfer presents a phase change to the AC controller that freaks it out, and it hard-stops the unit to avoid further damage. Carrier (who, BTW, market the entire generator/transfer switch package under their own brand) says my only option is to run the system in manual transfer mode.

Which brings me to my point -- if you're building your own controller, you might consider inserting a delay when transferring back to utility power, to give fussy controllers time to recognize a power failure and take appropriate measures. My AC controller, for example, won't restart for a short period after a recognized power failure. I'm pretty sure I can modify my ATS to accomplish this, but it's a low-priority item on my list right now.

alternety
10-06-2008, 09:56 AM
Thank you. I was aware of this. I kind of figured a commercial ATS would know about this. You can't do a simple switch back to the grid with some stuff. As you note, you need to open the generator link, wait, and then connect the grid. It can also damage electronics controlling motors.

There is a way around it but it is expensive. You make a whole house UPS. The inverters built to grid connect renewable energy sources sync their phase with the line. They have an internal ATS that can switch each way in milliseconds. But it means having expensive inverters and a battery bank. You could make the battery bank small and use your backup generator as a continuous charger. If you bought a bunch of battery, you could run the generator intermittently at mostly full power just when the batteries need charging. The idle/low load fuel consumption of a generator can add up during a long outage. This would probably save fuel; but I have no supporting data. You have a number of losses in this configuration. Rectifying and controlling the charging current, chemical conversion on both charge and discharge, inverter losses. If you have some sort of wind, water, or solar electricity, this is the way to go. Unfortunately I have no such resources available.

maintenanceguy
10-06-2008, 05:12 PM
I manage a dozen sites with back up generators. One of these has an ATS that "watches" the power utility sine wave when power is restored and waits until both the generator and POCO sine waves are in sync before transferring back to the normal (POCO) position. This typically takes anywhere from 5 to 20 minutes.

I'm not really sure of the advantage since it takes nearly 10 seconds to transfer to the generator during a power outage but at least the transfer back is smooth. (?)

Mikey
10-06-2008, 06:48 PM
I manage a dozen sites with back up generators. One of these has an ATS that "watches" the power utility sine wave when power is restored and waits until both the generator and POCO sine waves are in sync before transferring back to the normal (POCO) position. This typically takes anywhere from 5 to 20 minutes.

I'm not really sure of the advantage since it takes nearly 10 seconds to transfer to the generator during a power outage but at least the transfer back is smooth. (?)
I'd love to know the make & model of that switch. The advantage is you don't fry sensitive controllers, and continuous power is maintained. I can do without the latter, but the former is pretty important.

alternety
10-07-2008, 04:19 PM
Be careful about basing the transfer on phase congruity. You generator, if it is an inverter type like mine, probably does not vary enough to work properly. Most any mechanically controlled generator will probably drift; but no idea how quickly.