Power-Save Power Factor Correction Capacitors
Electrical Engineer Heinz Rosen recently contacted us in response to our new myth-busting efforts. He has been looking into power factor correction capacitors sold by Power-Save.
Original E-mail from Mr. Rosen: I've been looking into power factor correction capacitors sold by Power-Save and touted for residential use to save lots of money. They've recently gone into an advertising blitz on TV and mailings. You can find them on the web at www.Power-Save.com. To support their claims they prominently feature a University of Santa Clara Report written by a couple of Ph.D. Professors.
I had quite a few problems with how the product is used, the claims made for it and, in particular, with the University Report. I wrote to the President of the Power-Save (the only contact I was able to find) and questioned a number of assertions. Several of the University Report’s claims are questionable and some technical explanations dubious if not downright wrong, the explanation of the service connection, however, is scary in its ignorance and danger. Below is a copy of the letter.
March 12, 2007
Mr. Michael Forster,
President and CEO
Power-Save
3940-7 Broad St. #200
San Luis Obispo, CA 93401
Subject: Power-Save 1200 Questions
Dear Mr. Forster:
This letter is addressed to you because we were not able to locate any other customer contact individual to whom our questions could be directed.
Several families in our neighborhood saw the ad for Power-Save and being interested in saving money by reducing our energy expenses requested your brochure. After reading the brochure and the information provided on the Power-Save web page (http://www.power-save.com/), including the Santa Clara University report (Click here for report), certain questions arose. It would be appreciated if you could have someone provide answers to us before we commit to purchasing the Power-Save 1200 and having it installed in our homes.
For ease of reference the questions are listed below.
General Questions
1. To the best of our knowledge, based on information received from utilities, residential watt-hour meters measure KW-Hours not KVA-Hours. KW (dissipative) loads are independent of any reactive loads that may also be on the circuit. Since the homeowner only pays for KW-Hours how does improving the power factor reduce the KW-Hour charges to the homeowner?
2. Assume that the homeowner does pay for KVA-Hours. All inductive loads in a home are on only intermittently. The capacitive Power-Save 1200, however, is connected continuously. When few or none of the inductive loads of the home are on what prevents the power-factor from becoming negative (capacitive) and increasing the KVA demand from the utility compared to what it would be if the Power-Save were not connected at all during that time?
3. The brochure refers to the Power-Save 1200, the university report talks about the ABET 2201. How does one model relate to the other?
4. Power factor correction capacitors are normally rated in KVARs. There is no mention of the KVAR rating of the Power-Save 1200 either in the brochure or on the web. What is the KVAR rating of that model?
Questions Specific to the Santa Clara University report by Ayhan Mutlu, Ph.D. and Mahmud Rahman, Ph.D. Quotation marks surround sections quoted from the report. Our questions are in dark blue.
"3.1. Power Factor Correction. ABET-2201 brings the advantages of the power factor correction capacitor to household use. Figure 1 shows a typical connection of the ABET unit in a household. This unit is usually connected to the fuse panel where the electricity is distributed to different locations in the house. As mentioned above, the current passing through the current coil of the Energy-meter installed by the power distributor to monitor power consumption is the algebraic aggregate of the individual resistive, inductive, and capacitive currents flowing in different loads of the household, as shown in Fig. 1. Since the currents flowing in inductive and capacitive loads are half a cycle out of phase, it is possible to make their sum zero at any particular time by adjusting their magnitudes, consequently reducing the total current magnitude flowing through the Energy meter."
5. The authors seem to be unaware that most watt-hour meters measure just that, not the product of current and voltage. Although reducing power factor has beneficial effects for the utility by reducing line losses and potentially reducing distribution transformer sizes, there is no benefit to the rate-payer to improve power factor. What are we missing?
Fig. 1. An example ABET-2201 installation in a typical household.
"Due to the reduction in the total current, the power loss (I2total x R1) in the resistance R1, between the wattmeter and the ABET-2201, which varies from house to house, is also reduced. This is the instantaneous power saving that is achieved by installing the ABET-2201. It is important to note that i) the resistance R1 will depend on the locations of the Energy-meter and ABET-2201, and ii) the power saving is proportional to the square of the reduction in the current brought about by the ABET-2201.
Figure 4 presents average voltage, current, power, and power factor measured over a period of one hour in an actual household. It is clearly observed that the power consumption is reduced along with the improved power factor and reduced current."
6. Similar comment as in 5. However, the discussion concerning the power loss in R1 is relevant since it contributes to the KW load (paid for by the rate-payer.) See further discussion of this in question 7c.
7. There is no indication where the measurements represented in this figure were made. Presumably at the service connections after the energy meter shown in Fig.1. Nonetheless, some issues arise:
a. Since the inductive loads are intermittent and the power factor correction is continuous just what does “average” mean in connection with Q and power factor of Fig. 4? If it is the traditional meaning of the time integral of the function divided by the relevant period then it would not be applicable to the situation at hand. It is possible to choose a capacitive load that “on average” yields a power factor of 1.0 but since the energy measurement occurs continuously and the power factor alternates between some value between 0 and plus 1.0 and 0 and minus 1.0 there would be no or little reduction of the “average” current demanded from the utility, resulting in no or little savings to the rate-payer. Would you please comment on that?
b. From Figure 4 it might be deduced that the capacitive load of the ABET 2201 is implied at 1.347 KVAR (1773-426 VARS.) on the assumption that the inductive load is on continuously. However, this is not the case and when the inductive load is off there would be a 1.347 KVAR capacitive load, reducing the power factor and again increasing the current demand from the utility. The Figure(s) and text are unclear about exactly what measurements were made. Would you please comment on that?
c. Fig.4 shows a reduction of 36 W in actual power usage, attributing that result to a reduction of I2R losses in R1. Let us look at this more closely. An average single-family residence is generally provided with about a 150 A single-phase service. According to the NEC copper conductors for such a service are required to be AWG. No.1 minimum. The resistance of AWG No. 1 copper is 0.16 Ohms/1000 ft. The distance between the meter and the main panel rarely exceeds 50 ft. Thus the total distance for hot and neutral conductors is about 100 ft for a total resistance of 0.016 Ohms. A reduction of current from 10.66 A to 7.63 A calculates to a reduction of 0.146 W loss in R1. Where do the 36W come from? Once again the Figures and text are unclear. Would you please comment on that?
"Our experiments have shown that adding ABET-2201 has reduced the peak inrush current of the tested motor by about 5%. Also, when the motor was under full load, the unit has reduced the inrush current time about 15%."
8. Where were the motor inrush current measurements made? While it is true that a motor starting from stand-still has a high inrush current no theoretical reason is offered to explain why placing a capacitor across branch circuit terminals will reduce that inrush. Although a capacitive shunt at the input will reduce the transient seen on the supply conductors it is the capacitor (if it were large enough) that provides the energy for the inrush current which will remain the same as before. Let’s look at that a little closer. From the Fig. 4 implied 1.347 KVAR capacitive load it is simple to calculate the capacitance as 64μF. At 236 V this capacitor holds an average charge of 15x10-3 Coulombs. For a motor inrush current of 40 A. (Fig. 5) that amount of current is not even enough to sustain one mS of surge. We would appreciate an explanation of why the ABET–2201 reduces the motor inrush current.
Furthermore, Fig. 5 seems to show that even the running current of the motor is reduced when the ABET-2201 is installed. There is no mention of that in the text. We would appreciate an explanation of that, as well.
"This behavior of capacitors have lead to applications where these are connected in parallel with the power circuits of most electronic devices and larger systems (such as factories) to shunt away and conceal voltage and current fluctuations from the primary power source to provide a "clean" power supply for signal or control circuits. Such effects in capacitors can also be interpreted to act as a local reserve for the DC power source, …"
9.We don’t understand the comment about capacitors “…can also be interpreted to act as a local reserve for the DC power source…” What DC power source? If what is meant are the various DC power supplies integral to the various electronic equipment that may be in a home and that are connected to the branch circuits in the usual way, these contain their own built-in transient filters and surge suppressors. Could you please offer further explanation?
"The ABET-2201 in conjunction with the resistance of the household wiring forms a low-pass filter which prohibits higher frequency components from the incoming supply into household loads. Consequently, motors are subjected to less heating as illustrated in Fig. 7. "
10. While it is true that the 0.016 Ohm R1 and the 64μF ABET-2201 form a low pass filter we do not understand why the house wiring plays a role in that filter. Once the frequencies in question have reached the ABET they would propagate throughout the branch circuits unless there are other shunt capacitors (or series inductors) in the circuit. The house wiring would play a role if there is a capacitor directly across the motor but that would exist even without the ABET-2201. Thus we need to examine the low-pass filter formed by R1 and the ABET-2201. The “corner” frequency, f0, of that filter, that is the frequency at which the amplitude is reduced to half, can easily be calculated as 146 KHz! Every frequency below that will pass into the branch networks at nearly the same level as it enters the panel. Certainly the effect of that filter on the first several harmonics of 60 Hz. would be minimal.
Since there was no information provided in the report about the measurement of the harmonic content before and after installation of the ABET-2201 could you please provide us with some information concerning these measurements and a detailed explanation why the report shows the reduction in motor operating temperature and attributes this (without explanation) to a harmonic content reduction?
"4. The power distribution panel installed by distributor
The distributor of electrical energy installs a power distribution panel outside a household so that their personnel can monitor the total energy consumed by the household by reading the Energy meter installed in the panel. Two wires with 220 V across them are brought in from the locality supply grid into the distribution panel. All loads in the house which run on 220 V are connected across these two wires. Loads which run on 120 V are connected across one of these two wires and the ground terminal which is fabricated in each household by inserting a solid copper rod deep into the soil. All 120 V loads are divided into two circuits, each consisting of one the above mentioned two wires and the ground terminal. Figure 8 shows these two different configurations."
"Fig. 8. Distribution panel configurations in a household"
11. We had serious problems with this paragraph. To begin, the symbols used in Fig. 8 are totally different from the standard ones used in plans and there was no legend explaining the symbols, either in the text or in the figures. Therefore it was impossible to determine what is connected to what.
Single phase power distribution to residential homes is almost exclusively 120/240V provided by a center-tapped transformer secondary winding. The secondary winding carries 240V end-to-end while each end is 120V to the center tap. The center tap is the white neutral conductor that is connected to the neutral bus in the panel. The two black wires are connected to two power busses in the panel. Although grounding is a part of every safe distribution system it plays no part in routing the return current back to the source.
The text makes no explanation of how loads connected between the hot leads and ground rods without any additional connections or circuitry derive 120V from the 220V (sic) wires, the only ones that the text mentions as being brought from the utility. There is no mention of a return path. A distribution system such as described in the text would not work, would be unsafe, and would be in violation of the National Electric Code, that forbids using the earth as a return path.
We are surprised and baffled by this section.
We are prepared to purchase and install the Power-Save units once we have received satisfactory answers to our questions and look forward to your reply.
Send to a Friend