Combining multiple CTs and dealing with bi-directional flow 1

[abl33]

Our company is looking to expand a PV solar installation at our facility and because of accounting limitations with the local utility we may need to find a hardware solution to address what they cannot do in software.

At present, our single utility feed comes in to two 13.2kV switches which each have CTs that tie into two utility meters. At present, one of those transformers has solar panels on it and the utility attached a 2nd meter to track returned power generation.

We are looking to expand our solar production, but because of transformer sizing and other practical reasons, we need to install that power on the other transformer. We assumed that the utility could just put a 4th meter to capture the return power from the other transformer and have it all combined in the account, but apparently that is not possible, so my question is this... Is it possible to wire the CTs together to combine the usage into a single meter?

I know that this would mean that we would not be able to know how much power a specific transformer used, but that is not important to us at all.

From what I know, I believe that IF we have CTs on both transformers that are sized identically, that it is definitely possible to combine them and get valid readings at the meter, BUT I don't know if that is still true with bi-directional flow.

If I have 100A flowing in the "positive" direction on one set of CTs and 20A flowing in the "negative" direction on the other set of CTs, will that show up on the meter as 80A of "positive" flow? Likewise, if I have 75A flowing in the "negative" direction and I have 50A flowing in the "positive" direction, will that show as 25A of "negative" flow?

I really hope it would, but I could use a confirmation that this is in fact the case.

 

[waross]

The modern electronic meters track and record in separate registers internally:
Positive power.
Negative power.
Positive VARs.
Negative VARs.
Maximum demand since the last reset.
A programming set is required to set up and read the following.
Recorded at 15 Minute intervals:
Volts per phase.
Amps per phase.
Demand.
VARs.

One meter does it all.
The default programming is to sum arithmetically both positive KWHrs and negative KWHRs.
A common dodge to steal power used to be to mount the meter upside down for part of each billing cycle.
The electronic meters don't subtract KWHrs when the power is reversed, they continue to meter and record the usage.

The problem with metering reverse power may be an inability to program the meters.
The programming set is expensive.
If you are the only customer producing power they may not be able to justify the cost of a programming set.
If they are reading the meters remotely, there may be an issue reading more values from one meter.
But apparently they did it once for your other circuit.

There is no technical reason why the return power can not be metered.
They may use one set of CTs or two sets of CTs. (Their approved wiring diagrams may require two sets of CTs.)
The electro-mechanical meters were bi-directional.
Some tariffs did not allow credit for a leading power factor.
The KVARhrs were metered with a meter with a reverse ratchet that prevented th meter from turning when the PF went leading.
Theses were typically otherwise standard meters with a phase shifting transformer that made them responsive to VARs rather than KW.
Two of those meters would do the job.
One newer, electronic meter, properly programmed would do the job.
There is no technical reason why your plant can not be metered.
The utilities meter shop should be able to set it up with no problems at all.

Either the utility does not want you to increase your solar production,
or
The utility does not have the appropriate meters and or programming set in hand and does not want to incur the added expense.
or
You have been talking to a sales engineer who isn't very good at engineering.
I am not prejudiced against sales engineers, I have know some excellent sales engineers.
However not all are excellent.

Some good, legible photos of your existing meters would help.
Use the image tool above to post pictures.
ps: The cost of a programming set was mostly the charges for the proprietary software, not the hardware.
Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[abl33]

The issue isn't with the physical meters, but a limitation with the accounting software. They are able to aggregate multiple importing meters and combine them for billing. We currently have two meters that are handled this way. The problem is that apparently the software cannot aggregate multiple exporting meters to combine them for billing and this is what we would need to have happen to add additional solar to the site.

We currently have two 13.4kW siwtches and each have their own set of PT/CTs installed which each go to their own dedicated importing meter. There is not a 3rd set of PT/CTs for the exporting meter and it is just tied to the same feeds as the exporting meter. I don't think it is analog, but they both have "detents" to meter in only one direction. I know they use bi-directional meters for residential service, but their policy requires dedicated in/out meters for commercial scale. The existing system is ~300kW AC and we are looking to add another ~400kW AC. The local utility circuit apparently does not have a capacity problem and if we could just add it all to the existing meter it shouldn't be an issue, but adding it to that existing meter is totally impractical for a number of reasons such as the location and size of the 480V switchgear and the location and size of the 13.2-480v transformer.

Since I made this post, we have gotten a call setup with the chief engineer of the area next week to see what he says, but I was hoping to go into that meeting knowing if there is a technical reason why we would not be able to use two sets of CTs to generate valid results on a single meter. Our utility feed is a single feed to two adjacent MV switches. The PTs and CTs are located in those two cabinets and the outputs of those go to the meters which are all next to each-other. I was thinking that it would be ideal if all we had to do was replace one set of CTs so that both sets matched and then wire the CTs to remove one meter and replace it with a box and wire the CTs from that box into the existing box to combine them. I know this would be technically possible without problem if the loads were always both in the same direction, but what I don't know is if that would still be valid if one was exporting while the other was importing. I assume we could potentially move the PT/CT up in front of both switches if we had to purchase a 3rd switch, but obviously that would be much more expensive and require much more downtime than what I said above.

 

[waross]

The CTs may be connected in parallel to feed one meter.
The impedance of the CT secondaries is very high.
The impedance of a typical KWH meter current element is vary low.
The CTs are unable to force current through each other.
The current from both CTs passes through the KWH meter element.
You may suggest to the Engineer that he have his meter shop verify this.

An example of this is the use of delta connected CTs to meter unbalanced three phase power with a two element meter.
This is NOT a Blondel solution.
But the typical residential meter installation, with only a single potential element, on millions of homes in North America is not Blondel compliant.
(Unequal voltages will cause inaccuracies but the current metering is accurate.)

Another option may be primary metering ahead of both switches. The utility standards may preclude this solution.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[abl33]

Bill,
I know "normally", when the assumed flow of the current is in the same direction, that joining CTs this way provides good information for this purpose, but does is that still valid when you have changing directions of flow on either CT?

As for if it that or moving the CTs to an upstream location is allowed because of utility standards... that I"ll have to find out next week, but if this is a potential solution I would like to make sure they at least consider it. I've dealt with the utility engineer in the past and he is a pretty reasonable guy as long as it is safe and legal.

I do assume worst case we could install another MV disconnect in front of our existing as we own the existing, but as stated previously, that is more of a backup plan.

 

[argotier]

Very interesting.

May I ask if the existing CT/PT are connected to the MV or the LV side?

Energy meters take into account not only the currents but the voltages, if the PTs are connected on LV side of the transformers and you have to choose one for the exporting meter you may introduce some error, as the voltage drop on each transformer will depend solely on their respective load. If PTs are on MV side this is not a problem.

I know "normally", when the assumed flow of the current is in the same direction, that joining CTs this way provides good information for this purpose, but does is that still valid when you have changing directions of flow on either CT?

By connecting both CTs in parallel you will combine the currents and should give you a proper reading. But consider that the currents will add aritmethically only if both cos φ matches, generally they will sum geometrically (taking the current phasors).

I know that this would mean that we would not be able to know how much power a specific transformer used, but that is not important to us at all.

You wont lose the ability to measure each transformer power, as the paralleling of the two CTs should be done "downstream" the importing meters, the only meter that will be feeded with the combined currents should be the exporting one (this could be more clear with a drawing, but my attempt at it was pretty lame).

So far, is theoretically possible, but I see one problem:

What happens if one of the MV switches is opened? The de-energized CT wouldnt act as a load for the energized one, drawing a little of current away from the exporting meter?

 

[abl33]

The existing CTs and PTs are connected to the MV side of our equipment. they are physically inside the MV disconnect switches on the side of the switch between the switch and the transformer. The meters use the PTs for voltage to convert the 13.2kV to 120V for the meter and use the CTs as is. I haven't measured it, but I can't imagine the voltage differential on either of these switches to be anything meaningful even if one side is under heavy load and the other side isn't.

I get what you say in theory that there could be a situation with an induced inaccuracy if once side is turned off, but in reality, neither of these switches would ever be turned off for any meaningful time unless there is some major electrical disaster on-site like a transformer failure. Its good to be aware of that, but I do not expect the utility to seriously consider that an issue. In the 20 almost years we have been in the facility the only time the MV switch was turned off what after one of the PTs shorted and failed which caused use to lose some of those MV fuses (BTW, those are pretty cool to see armed/installed with those silencers).

The phase angle/power factor issue is I guess what I am potentially concerned about. Our PF under normal load is 0.92-0.98 (our major loads are all resistance loads), but I assume that when we are exporting power the power started to go leading.

 

[bacon4life]

Although CTs used to be routinely paralleled, it seems like a bad solution for a brand new installation. Paralleled CTs work fine in normal situations, but have some traps when it comes time for maintenance or meter testing. Adding an oddball setup creates risks for the metering technicians.

It may be possible to tie two physical meters together using pulse hardware. Do know the models of your existing 3 meters?

One of the drawbacks to AMI/AMR is that metering folks can have a lot less flexibility in programming meters specifically for one-off situations. I can see it costing as much to update the utility billing system as it will cost to install the solar panels.