Backfeeding PV thru air conditioner circuit

[noahbaude]

HVAC PV add on energy reduction appliance at commercial strip mall.

Here's the design for each rooftop air conditioning system. Please comment on what could possibly go wrong or what I'm missing.
(6) 350W 48VDC panels installed on flat roof using ballast mounts
(3) DS3 APS microinverters attached under every 2 panels
(1) APS ECU monitor
microinverter single phase 240VAC power outputs wired to load side disconnect of 3-5 Ton air conditioning condensers or package units in sight of panels.
Not looking for utility electricity buyback - just reduction in local energy use especially during summer peak periods.

Cheers,
nb

 

[ggunn]

HVAC PV add on energy reduction appliance at commercial strip mall.

Here's the design for each rooftop air conditioning system. Please comment on what could possibly go wrong or what I'm missing.
(6) 350W 48VDC panels installed on flat roof using ballast mounts
(3) DS3 APS microinverters attached under every 2 panels
(1) APS ECU monitor
microinverter single phase 240VAC power outputs wired to load side disconnect of 3-5 Ton air conditioning condensers or package units in sight of panels.
Not looking for utility electricity buyback - just reduction in local energy use especially during summer peak periods.

Cheers,
nb

It sounds a little flakey to me; will you be complying with 705.12 in whatever feeds the HVAC disco?

 

[ggunn]

It sounds a little flakey to me; will you be complying with 705.12 in whatever feeds the HVAC disco?

Also, it sounds like you will be bypassing whatever OCP is protecting the HVAC unit. I think you need to connect at the MDP rather than out on that HVAC disco.

 

[noahbaude]

After reviewing 705.12 it looks like we'll need fuses or circuit breaker on the micro inverter conductors before the disconnect. Plenty of room in the AC control panel. Overcurrent protection on the HVAC unit would remain the existing CB at the MDP.

(1) Dedicated Overcurrent and Disconnect. Each source interconnection of one or more power sources installed in one system shall be made at a dedicated circuit breaker or fusible disconnecting means.

 

[ggunn]

After reviewing 705.12 it looks like we'll need fuses or circuit breaker on the micro inverter conductors before the disconnect. Plenty of room in the AC control panel. Overcurrent protection on the HVAC unit would remain the existing CB at the MDP.

(1) Dedicated Overcurrent and Disconnect. Each source interconnection of one or more power sources installed in one system shall be made at a dedicated circuit breaker or fusible disconnecting means.

If you are adding PV current on the load side of the HVAC disconnect you are defeating the OCPD device for the HVAC and you may be overloading the conductors. Also, have you considered 705.12 for every busbar and feeder all the way back to the utility meter? I could be misinterpreting what you are saying, though; can you draw a one line of your circuits?

 

[winnie]

If the goal is to reduce HVAC consumption with a guarantee of zero export, how are you arranging to prevent export when the compressor is not in use?

 

[mtnelect]

If you have a Photovoltaic System and you turn on your HVAC, it will suck the life out of it. Please see the attachment CEC for future workshops to deal with a solution for the above.

 

[pv_n00b]

It's possible. There were a few big names in the HVAC industry who tried to make this work, I don't know if any are still doing it though. A few issues:

• You will need an OCPD between the PV and the HVAC disconnect
• The conductor from the disconnect to the RTU will fall under 705.12(B)(1) and the conductor is probably not rated for both the HVAC OCPD and the PV
• The HVAC RTU probably has a max OCPD protection rating that is already taken by the HVAC OCPD. No headroom to count the PV contribution.
• Even if you don't expect to export you have to design for it unless you are installing protection to prevent any export
• All the conductors, bus bars, and OCPD from the HVAC disconnect to the service will have to be reviewed for compliance 690 and 705.12

 

[wwhitney]

• The conductor from the disconnect to the RTU will fall under 705.12(B)(1) and the conductor is probably not rated for both the HVAC OCPD and the PV
• The HVAC RTU probably has a max OCPD protection rating that is already taken by the HVAC OCPD. No headroom to count the PV contribution.

Couldn't both of these issues be dealt with by connecting the line side (grid side) of the PV disconnect to the line side of a fused HVAC disconnect?

Cheers, Wayne

 

[jaggedben]

Wayne's idea effectively turns the existing circuit into a feeder and solves the problem for that circuit. I was going to suggest replacing each existing disconnect with a small subpanel with one breaker for the HVAC and another for the PV. It amounts to the same thing.

But none of that solves the questions upstream about complying with 705 in whatever panels are feeding the existing circuits.

If we were talking about one such system and the load at the meter never fell below the 1.5kw or so we are talking about, I'd say the main problems are legal and one could probably get away with it if no one is looking closely. (But since anyone will eventually be able to see the panels in Google maps, is it worth the risk of penalties?) But in any case the OP said for there'd be one such system for 'each' HVAC unit, and didn't say how many. Which makes me further question whether why it shouldn't just be one large system with a single connection and interconnected through the proper process.

 

[jxofaltrds]

I just watched this yesterday. Hope it helps:

Could be this one. I watch so many:

 

[ggunn]

But in any case the OP said for there'd be one such system for 'each' HVAC unit, and didn't say how many. Which makes me further question whether why it shouldn't just be one large system with a single connection and interconnected through the proper process.

Many people who are new to grid tied solar and/or do not understand how it works erroneously believe that a PV system can be used to offset a specific load and that it matters where in their electrical system it is interconnected. Maybe that is what is going on here.

 

[winnie]

I think post 12 has it.

To the OP, if you connect the output of the micro inverters anywhere to the AC grid, at the main panel, at a subpanel, or at a branch circuit, then your PV will feed the grid.

If your goal is zero backfeed to avoid the various interconnection hassles, then you will need to rethink things.

There are systems which can use current sensors to guarantee 'zero backfeed'. I don't know if these are a legal solution to avoiding interconnection requirements.

There are specific HVAC systems designed with direct PV connections which guarantee zero backfeed.

You might be able to arrange some sort of interlinked contactor so that the PV system is only connected when the compressor is drawing power. Such a system would probably not be easy to get approved.

Jon

 

[pv_n00b]

Couldn't both of these issues be dealt with by connecting the line side (grid side) of the PV disconnect to the line side of a fused HVAC disconnect?

Cheers, Wayne

That will do it for the downstream. Assuming the fused disconnect itself can handle the total feed.
While the OP did not state it I get the feeling they are wanting to avoid the utility interconnection process by interconnecting deep in the system close to a large load. That's not an option.

 

[ggunn]

That will do it for the downstream. Assuming the fused disconnect itself can handle the total feed.
While the OP did not state it I get the feeling they are wanting to avoid the utility interconnection process by interconnecting deep in the system close to a large load. That's not an option.

Interconnecting PV with the HVAC supply conductors looks like a bad idea to me no matter how you slice it.

 

[winnie]

IMHO if the PV is coupled to the HVAC system in such a fashion that you reduce the cost of running the HVAC system without incurring the costs and complexity of interconnection, then it might be a good idea. You give up any benefit of using excess PV to supply any other loads (or feed back to the grid), but if the HVAC system runs for a large fraction of the time and the PV is smaller than the HVAC load, it might pay.

I know that you can buy minisplit systems designed to be directly coupled to PV panels. This greatly reduces the 'balance of system' costs associated with the panels and eliminates any interconnection concerns. But this is a _specific_ mini split HVAC unit designed to connect to PV, not a generic HVAC system someone is trying to retrofit. Also: this is a more expensive way to buy a minisplit, and I don't know if the system properly addresses things such as rapid shutdown.

-Jon

 

[pv_n00b]

IMHO if the PV is coupled to the HVAC system in such a fashion that you reduce the cost of running the HVAC system without incurring the costs and complexity of interconnection, then it might be a good idea. You give up any benefit of using excess PV to supply any other loads (or feed back to the grid), but if the HVAC system runs for a large fraction of the time and the PV is smaller than the HVAC load, it might pay.

I know that you can buy minisplit systems designed to be directly coupled to PV panels. This greatly reduces the 'balance of system' costs associated with the panels and eliminates any interconnection concerns. But this is a _specific_ mini split HVAC unit designed to connect to PV, not a generic HVAC system someone is trying to retrofit. Also: this is a more expensive way to buy a minisplit, and I don't know if the system properly addresses things such as rapid shutdown.

-Jon

I agree that it would be a good idea. But any PV system connected to the AC side could export unless there was some system in place to monitor the POCC and disconnect the PV system if it started to export due to low load at the site. And the reverse power flow systems that a utility would accept are too expensive for it to work out. Now a DC-coupled PV system directly to the HVAC system would not back feed so that should work. But the PV system still would need to be code compliant with RSD and everything else and the HVAC system would have to be designed for PV. So that would probably run up the cost.
I guess it all comes down to the cost is too high to make this work.