The electrical engineering department at the Public Works Commission for the City of Fayetteville (Fayetteville, North Carolina) recently encountered an unusual problem at a metering installation. The customer had two 400-hp motors, and when one motor ran, the meter ran forward. When the other motor ran, the other metered service ran backwards.

One 2500-kVA substation-type transformer with internally mounted current transformers (CTs) served the electrical service. The primary distribution voltage was 12,470 V grounded-wye and the secondary voltage was 4160 grounded-wye. Under normal metering applications, the CT polarity side is on the line side of the service.

Many technicians use this general rule of thumb: “If the meter runs backwards, switch the currents or the voltages, but not both.” Sooner or later, however, this won't work.

The service fed two 400-hp motors that the customer alternated using 24 hours per day, seven days per week. As previously stated, the problem was that the meter ran forward on one motor but when the customer switched to the other motor, the meter ran backwards. The meter that was employed at the service was a Type 9S 4-wire wye, which fed a data recorder. One of the line crews verified that the customer was hooked up correctly and was serving the motors correctly.

Investigating the Problem

Our engineers met the metering crew, along with one of the line supervisors, at the site to measure the voltage and current. We made sure the equipment was providing the proper voltage and using the correct current to supply a 400-hp motor. We determined that the internally mounted CTs were set on the 200:5 tap, or a 40 multiplier, and the voltage transformers were at 2400:120, or a 20 multiplier. We hooked the phase-angle meter to measure the angles of the phase voltages. We observed B_Phase about 120 degrees from A_Phase and C at 240 degrees from A_Phase, which is about a 120-degree separation as it should be. We also did this for the currents and got about the same 120-degree separation. We then measured the phase angle between A_Phase voltage (A_V) and A_Phase current (A_I) and got 87 degrees, B_Phase voltage (B_V) to B_Phase current (B_I) and got 84 degrees, and C_Phase voltage (C_V) to C_Phase current (C_I) and got 88 degrees. The meter was running forward at this time. We then drew a phasor diagram and started to scratch our heads because this did not make sense (Fig. 1).

We then reversed the current lead on the phase angle meter and observed AΦ - 267 degrees, BΦ - 264 degrees and CΦ - 268 degrees. This didn't tell us anything that we recognized. We then decided to check the leads coming from the CTs and Pts and discovered that B_V was actually connected to A_I, C_V to B_I and A_V to C_I. We then reconnected the voltage leads correctly and again performed phase-angle readings. We observed AΦ, - 150 degrees, BΦ - 150 degrees and CΦ - 150 degrees. This still didn't make sense, since we were fairly sure that the phase angle should be about 30 degrees. We then reversed the phase-angle meter leads and got each phase angle to be 30 degrees. This told us that either the whole CT was mounted backwards or some of the CT secondary leads were wired backwards. We first verified that the secondary CT polarity was wired correctly. The metering crew then wired the meter using the reverse polarity from the CT. After four hours of digging, the meter ran correctly and was measuring the proper watts at the proper phase angle (Fig. 2).

Examining Other Locations

As we were leaving the site, the meter crew mentioned they had two more identical sites similar to this existing installation. When we got back to the office, we looked at the difference in what we were charging the customer and what we should have been charging the customer. Using the 87-, 84- and 88-degree angles, and comparing those to using the 30-degree angles, we determined that we were charging the customer only 7% of the total we should have been charging (Fig. 3).

A few days later, we met the metering crew and the line supervisor at one of the local hospitals with a similar meter on one transformer and another meter on two identical substation-type transformers. The first meter we checked had the same problem. The voltage leads had been switched to make the meter run forward. The real problem, however, was that the CT polarity was reversed. This had about the same results as to the amount of revenue we had lost as the first location.

At the second location, we discovered that instead of switching the voltage leads from the voltage transformers, the leads from the CTs on each transformer were switched. This gave the overall phase angles of 335 degrees, instead of the actual phase angle of 36 degrees. By aligning the current leads and using the reverse polarity, a phase angle of 335 degrees with a power factor of 90% became a phase angle of 36 degrees with an actual power factor of 80%. This meant that we had been over-billing the customer about 10% since the transformers were installed about 15 years ago. The metering crew rearranged the CT leads to the proper phase voltage and to the reverse polarity so that the meter was reading correctly.

Mounting Transformers Backwards

Our Transformer Shop removed the hand-hole cover from one of two spares we had on the yard. As we had suspected, the CTs, which were made by the same manufacturer, were mounted backwards. PWC has 11 of these transformers in service. Fortunately, all but these four transformers were behind primary metering, and the CTs mounted backwards don't have any effect on the revenue we collect.

We were trying to convince ourselves that the manufacturers knew how to hook up a metering CT on the secondary side of the transformer to give polarity as specified in IEEE C57.13. It made no sense to PWC that a manufacturer would intentionally install a CT backwards. The manufacturers responded that CTs are installed in this manner for the benefit of differential relaying applications.

We visited a substation-type transformer manufacturing plant. After talking with their engineering and testing crew leaders, it still did not make sense that the manufacturer would install CTs for differential relaying applications when the specifications had called for a metering application.

We also questioned why the manufacturer would install metering accuracy CTs the same way it installed relay accuracy CTs feeding a differential relaying application. For that matter, we also didn't know why the manufacturer would install low-side CTs backwards for differential relaying applications (Fig. 4).

The metering crew informed PWC that they had been instructed to switch leads until the meter moved in a forward direction, never having the concept of phase angles explained to them. Many utilities use that same approach in transformer-rated metering applications.

The Polarity Dot

PWC Engineering recognizes changing ANSI standards can cause too much change. This brings us to the polarity dot, also known as the Magic Dot. Think of the great chaos in the industry if some transformers came wired with dots out, while others came with dots in on their CTs, depending on whether the customer might want to use the CTs for metering or for differentials, as suggested by Chuck Whitley, utility director emeritus from the city of Wilson, North Carolina, who says to focus on the issue, not on the dot.

PWC Engineering recommends reading and understanding the substation-type transformer nameplate. Utilities should focus on the application of the transformer and be prepared to employ reverse polarity CT output for metering purposes.

Consider the way manufacturers wire differential relay CTs. One manufacturer installed the CTs on the high-voltage side to orient the polarity dot facing the incoming leads and on the low-voltage side with the CTs installed with the polarity dot facing the outgoing leads. All manufacturers use this method for differential relay applications. To avoid confusion, the manufacturers will install the CTs in the same manner for metering applications. This is intended to avoid confusion.

As we investigated the use of the low-side CTs, we realized that these CTs were used showing the current coming out of the reverse polarity contact. We considered turning the CT around so the line current went into the polarity dot to give positive polarity from the proper contact. All CTs could be mounted in the same fashion with the other CT mounted on the low side of the transformer. The metering CTs and the hot-spot CTs would be mounted the same way, giving positive polarity from the proper contact on the low side of the CT.

PWC Engineering consulted with Wayne Hartmann with GE Multilin to look at the feasibility of turning the low-voltage side CTs around. In his opinion, it was far easier to accommodate the current flow direction presented to the differential relay from the low-voltage side CTs. During load flow, the currents presented to the high-side and low-side inputs of the differential relay would be 180 degrees out of phase.

Changing the Magic Dot may create too much confusion. The best thing we learned was to verify all 3-phase metering installations with a phase-angle meter, which will quickly pay for itself. PWC Engineering plans to do just that from this point forward.

---

Rick Anderson is chief electrical engineer and manager of electrical engineering for the Public Works Commission for the City of Fayetteville, North Carolina. He has has more than 28 years of experience in the electric utility industry and is a registered professional engineer.
rick.anderson@faypwc.com

Tommy Cooper is the materials standards engineer for the Public Works Commission for the City of Fayetteville, North Carolina. Cooper has more than 40 years of experience in the electric utility industry and is a registered professional engineer.
tommy.cooper@faypwc.com

Lessons Learned

• If you purchase substation-type transformers with internal metering current transformers (CTs), equip your metering crews with a 3-phase, phase-angle meter. Properly train the metering crews and explain why they need to use this type of tool.

• If you specify and purchase substation-type transformers with internally mounted current transformers, ensure that the CT secondary polarity is positive and on the lowest-numbered connected output according to IEEE C57.13.

• Check out all of your substation-type power transformer installations and verify the nameplate information.

PWC Engineering's plans include proposing a change to transformer testing standards that states “a manufacturer must verify the polarity of any internally mounted current transformer outputs.” PWC Engineering believes that the low-side CTs, mounted with the H₁ polarity dot facing the low-side bushings, which gives a reverse polarity output from the CTs, should be mounted with the polarity dot facing the transformer winding which would give a positive polarity from the lowest-numbered output from the CT as stated in C57.13.