Any way to improve power factor in off grid solar system

pmarchio

Hi, I've looked several places, but can't seem to find any info about improving the power factor in a small solar set up.

I built an off grid solar system in my rv with 450 watts, 4 t105 batteries, and a cobra 2500 watt inverter. I recently bought a 26" led that claimed to consume only 21 watts, and .2 watts on standby. But when I plugged it in and looked at my kill-a-watt meter it was using 3.6 to 4a instead of the anticipated 1.7a, the inverter has an 88% efficiency, but I figured that couldn't account for the difference. That's when I started to learn about true and apparent power. My understanding is that the tv only uses 21watts in real power, but is consuming close to 48 in reality. I'm looking online to see if there's anything I can buy or do to improve the power factor and I'm wondering if there is anything or if I need to just buy an energy efficient tv that uses 12v dc.

Thanks for any help ahead of time.

Paul

NorthGuy

Re: Any way to improve power factor in off grid solar system

Low power factor caused by LEDs (unlike motors) might be as low as 0.5 and is hard to improve. But it shouldn't be a concert because you consume real power, not the apparent power. Kill-a-Watt should be able to measure both - use W/VA button. 21W LEDs should show as 21W and up to 40VA.

pmarchio

Re: Any way to improve power factor in off grid solar system

Ok, great that's the info I needed, I assumed that the tv only needed 21w to run, but that it consumed ~48w in the process. But your saying that it is only taking 21w from the batteries.

Yes, the kill-a-watt has apparent and real, that's in fact how I started to learn about it. Yes, it shows just under 21w and about 45VA. Its just the amps show .4a (@ 113V) so I assumed its consuming ~45w

inetdog

Re: Any way to improve power factor in off grid solar system

pmarchio wrote: »

Ok, great that's the info I needed, I assumed that the tv only needed 21w to run, but that it consumed ~48w in the process. But your saying that it is only taking 21w from the batteries.

That is generally correct, but the details go a little deeper than that.

The inverter must deliver the current consumed by the load, whether that current is transferring power to the load or not. As long as the inverter has a way to take the energy which is being returned as part of the apparent energy cycle, then that power will not have to come from the batteries. That can be done either by storing energy internally in the inverter over the period of one cycle or actively returning power to the batteries part of the time.
Even if this is done perfectly, there will be higher resistive losses and higher Peukert's Law losses in the wiring and the battery caused by the difference between the apparent power and the real power.

There is not much that you can do to correct for the distortion (non-linear) power factor which comes from inexpensive LEDs or CFLs and from inexpensive switching power supplies. What you can do is select light fixtures and bulbs and DC power supplies that present a reasonably high power factor and also look for inverters that handle a low power factor well.
You need to size your wiring based on the full current corresponding to the apparent power, not just the real power.

And you can use your Kill-a-Watt in combination with a DC ammeter to see just how high or low power factor affects the power that your inverter pulls from the batteries. There is no such thing as low power factor for constant voltage DC, so any current coming from the batteries represents real power taken from the batteries.

vtmaps

Re: Any way to improve power factor in off grid solar system

inetdog wrote: »

As long as the inverter has a way to take the energy which is being returned as part of the apparent energy cycle, then that power will not have to come from the batteries. That can be done either by storing energy internally in the inverter over the period of one cycle or actively returning power to the batteries part of the time.
<snip>
And you can use your Kill-a-Watt in combination with a DC ammeter to see just how high or low power factor affects the power that your inverter pulls from the batteries. There is no such thing as low power factor for constant voltage DC, so any current coming from the batteries represents real power taken from the batteries.

inetdog, it will take a specialized DC ammeter to pick up the current spikes into the battery if the inverter is returning power to the batteries during parts of each cycle. Do any of the common inverters actually work that way?

--vtMaps

NorthGuy

Re: Any way to improve power factor in off grid solar system

inetdog wrote: »

The inverter must deliver the current consumed by the load, whether that current is transferring power to the load or not. As long as the inverter has a way to take the energy which is being returned as part of the apparent energy cycle, then that power will not have to come from the batteries.

Low power factor for LEDs, CFL Bulbs, and electronics results from the power being drawn at the peaks, when voltage is high, and not drawn when voltage is low. The power never flows back to the inverter as would be the case with inductive loads.

You're certainly right about resistive losses being higher, but if the wiring is done right, this is negligible.

bill von novak

Re: Any way to improve power factor in off grid solar system

vtmaps wrote: »

inetdog, it will take a specialized DC ammeter to pick up the current spikes into the battery if the inverter is returning power to the batteries during parts of each cycle.

An oscilloscope with a current probe is a fairly easy way to make this measurement. Many scopes nowadays will take modules that allow power factor analysis, which is very helpful.

Do any of the common inverters actually work that way?

Most do. The output bridge is on for half the AC output cycle; if you are feeding current back at any point in the cycle, the bridge switches pass it back to the battery as easily as they pass it out of the battery.

Also keep in mind that there are two kinds of bad power factor:

1) Leading/lagging. Capacitive loads "lead" - they take current earlier in the cycle than a pure resistive load would. Inductive loads "lag" - they take current later. Inductive loads are the most common since large AC motors present inductive loads; inductive loads can be compensated for by adding capacitance.

2) Harmonic distortion. Cheap power supplies and LED bulbs do something called "peak rectification" where they take power only during the very peak of the cycle. That means a 1A (average) load might take zero current for most of the cycle then 3A at the peak of the cycle. This puts a higher load on the inverter. The only way around this is to filter the load with a large resonant filter. These are generally too large to be practical for most users. As time goes on, regulations on harmonic distortions have been forcing power supply manufacturers to go with better power factor designs, fixing a lot of these problems.

BB.

Re: Any way to improve power factor in off grid solar system

And remember it is not 20 or 40 Watts. It is 20 Watts and 40 VA.

Watts = Volts x Amps x Power Factor

VA is the "raw" measurement and Watts is the derived measurement.

Wiring, generators, fuses/breakers, inverters, and such care about VA on the AC electrical side.

Generator fuel and battery usage care about Watts... It does get more complex, but those are the takeaway basics.

-Bill

NorthGuy

Re: Any way to improve power factor in off grid solar system

vtmaps wrote: »

inetdog, it will take a specialized DC ammeter to pick up the current spikes into the battery if the inverter is returning power to the batteries during parts of each cycle. Do any of the common inverters actually work that way?

Here is typical AC current draw that I get from my battery manager. It is computers, lights, electronics. You can see that it draws on top of the peaks and there's almost no consumption in between.



Here's DC draw from the batteries. It does go below 0, so inverter does return some power to the batteries, but the waveform doesn't seem to resemble AC draw an any way. I guess Inverter has lots of capacitors, inductors, and a big transformer, so the power gets heavily shifted between the cycles.



On the pictures above the black line is an average of about 150 cycles at the frequency stated on top of the picture. The blue lines are maximum and minimum of individual measurements. On the bottom picture you can see some clipping on the bottom. This is because this is a one-directional meter, the opposite current is measured by a complementary meter connected in opposite direction. Of course, Inverter supplies both L1 and L2, while the AC picture is of L2 only, but L1 looks very similar.

BB.

Re: Any way to improve power factor in off grid solar system

Technically, the Inverter DC input current should be a "sine squared" wave form (the instantanious DC current needs to supply the instanous AC power -- i.e., P=I²R).

Of course, if the AC loads are not sine wave voltage/current, the DC input current will not be a sine sqaured wave from either.

So, this is the case in DC where the DC Average value is not the same as the RMS (root mean square) value of the DC current (i.e., there is a "power factor" in the DC equation too).

If there was enough capacitance/energy storage in the AC Inverter to "smooth out" the DC current, you would not see that "spikey" DC current wave form.

For computers, which can change DC current requirements on the nano-second level, this actually cause us some grief when designing Power Factor Corrected AC power Supplies.. If there were large enough swings in the computer DC power requirements, it would cause low power factor on the AC input side of the supply (i.e., the AC supply was ~1.0 PF corrected, but the changes in current flow during a few AC input cycles were enough to make it fail the PFC correction specifications).

There are a whole lot of layers in this onion when desiging power systems.

-Bill

pmarchio

Re: Any way to improve power factor in off grid solar system

Thanks for a the further information as well. I spent some of last night looking things up and reading basics on AC power, as when I put things together, I probably focused 95% of my attention on the DC side. I feel like my dc wiring to the inverter is acceptable as my largest load is ~900w (microwave) and I'm using 3.2 ft of 1/0 awg to the inverter and battery interconnects are all 1/0 8" long (4 t105s in parallel series).

But on the AC side I didn't really give much consideration. I have the inverter plugged directly into the rv's shore power cable (w converter unplugged) using a 15a pigtail adapter in order to provide AC power to all the outlets. When I took the readings from the kill-o-watt, I forgot that my broadband wireless wifi was still plugged in, but it only uses .04a @112v. But I unplugged that, and then compared the the tv's amps watts and va with the tv in standby and powered on, both though the Rvs AC wiring, and then just plugged directly into the inverter itself. The rv seems to have a .03a @112v ghost load that I can't find anywhere, that it seems to consume, I unplugged everything I could find including unfilled power strips, but .03a seemed to be its baseline, as compared to 0.0a when plugged directly into the inverter. But by and large the results seemed pretty similar, the numbers would fluctuate, but they were about the same with the rv side seeming ever so higher. The one thing that seemed interesting to me was the difference between standby mode and power on. In standby mode the kilowatt meter would fluctuate between 0va and 40va, sometimes landing in between(20va), 0w and 2.1w, 0a and .37a (@ 112v, 61.0 hz), with the power factor switching between 1.0 and .02, whereas with power on it showed 39va, 16w, and .35a with a pf of .40. It seems like the tv requires the same amount of energy to be present whether on or off, but just consumes more when on...

Then finally, I found some other info online that supported the idea that the inverter would not consume the reactive power from a non linear device like LEDs, and that the kilowatt meter is designed for on grid systems and therefore it's readings aren't accurate as far as what's being actually consumed in an off grid system except for Watts.

BB.

Re: Any way to improve power factor in off grid solar system

It is hard to tell with electronics these days... "Standby" can mean all sorts of things--From being virtually "off" to being pretty much "on" running background tasks (talking with the satellite/updating program guides, recording shows to hard drive, etc.).

Consuming power vs not consuming power... This gets into physics and engineering definitions. You can have "systems" that do no work. For example, if you have a spring and push agasint it, then pull back--You temporally did work compressing the spring, but it returned the power when you pulled your hand back--So there was "no net work performed".

With PF=0.0 -- It is that sort of system where there is (virtually) no net work being done--But you still need to supply the current/voltage so the device can turn on with the remote control (for example).

And, the Kill-a-Watt meter is not that accurate--But it is usually "good enough" to get a system designed. You are never going to run your inverter at 100% of rated power but closer to 50-75% maximum of rated power (in most cases). Digital instruments sometimes do a good job of "lying" to us with lots of digits and pretty repeatable results--However, their accuracy is sometimes not the greatest (Kill-a-Watt meters are much better on Sine Wave/Utility power vs running on Square Wave AC inverters and some small gensets).

That is why a basic Fluke true RMS reading DMM may cost around $300 minimum. It is not only the name they are charging for.

-Bill

NorthGuy

Re: Any way to improve power factor in off grid solar system

pmarchio wrote: »

In standby mode the kilowatt meter would fluctuate between 0va and 40va, sometimes landing in between(20va), 0w and 2.1w, 0a and .37a (@ 112v, 61.0 hz), with the power factor switching between 1.0 and .02, whereas with power on it showed 39va, 16w, and .35a with a pf of .40. It seems like the tv requires the same amount of energy to be present whether on or off, but just consumes more when on...

Looks like in standby mode, it is pulsing on and off. Probably doesn't want to be on all the time to save energy, but still needs power to wait for a remote.

pmarchio wrote: »

... kilowatt meter is designed for on grid systems and therefore it's readings aren't accurate as far as what's being actually consumed in an off grid system except for Watts.

If you have an MSW (Modified Sine Wave) inverter, some meters can fail to measure correctly. If you have PSW (Pure Sine Wave) inverter, then there's no difference with the grid.