Being a large industrial electrical consumer at work we, like most others, have to pay a penalty per Kwh if our power factor falls below a certain point. Last year it was 80%, at the last Corporation Commisssion rate meeting they pushed for 90% but got 85% and may get 90% this year so it's something I've got to watch.

When we first moved into the building it was hovering in the low 80s, I managed to get that up to 98% during last winter by utilizing and moving around some of the PFC units that were left in the building when the previous tenent moved out but of course we weren't running the mainchiller that provides all the air conditioning. As we went into this summer it started dropping and at last billing it was down to 92%. Nothing has changed in our normal everyday load situation so that only left the chiller as upsetting the apple cart.

Today I finally got in a set of PFC caps I had to order from GE, one to put across each of the four compressors in the chiller. I went separate as the chiller will vary the number of compressors it runs according to the load it's seeing thus they only come into play if the contactor is active for any particular compressor.

While it will be another couple of months until I see the actual numbers stabilize on the billing I do know that the chiller, which was drawing 106.7A (480V 3 phase) under full load has now dropped to only pulling 100.6A - a 5.47% reduction. So long as we're still over 85% we'll get billed the same per Kwh at 85% or 100% PF, but the reduction in amperage draw can only mean that the motors are wasting less power due to true vs apparent. Also, by needing to pull less amperage through the wiring, the voltage drop along the lines will be less, my chiller compressors will see a bit higher voltage and will operate more efficently.

While there's no real payback involved for a common residential customer since very few are monitored for PF percentage it's going to make a small reduction in our electric bill and help protect me from penalties should they be granted the 90% number this year.

Point is when operating a large AC motor, even single phase, it pays to keep the reactive power losses to a minimum.

 

Yes the current will pass through a minima as the cap VAR increases and PF gets closest to PF, before it heads off into capacitive loading. Definitley worth placing the cap with the load, and behind the contactor, otherwise it may turn counter-productive. It must be a bit frustrating not to have a meter and having to wait for bills to come in that may blur the picture for individual changes you make.

Getting up to 98% means you must have pretty low harmonic levels with all your loads.

 

Without a true KVAR meter for 480V it's been hard - lots of research, math and measurements with clamp-on amp meters and good voltmeters mixed in with good old experimentation.

Why that chiller didn't come with a PFC option or at least suggested values is beyond me. I even snaked through a lot of people to get to the proper engineering department at the manufacturer and they had no suggestions as the question had never arisen. Sometime in the next few days the head engineer I was dealing with at the last is going to get a long letter because, in my opinion, the individual compressors should exhibit a decently longer amount of life and relaiability since there's obviously less strain on the motors.

I'd say our load breaks down as 25% motors, 25% T-8 fluorescent lighting, 45% pulse-start MH high-bay fixtures and 5% general office equipment. I've also got quite a few 480Y/277 to 208Y-120 step-down transformers but they're only seeing minimal usage since we don't have any 208Y equipment like the previous occupant did.

I want to lose the MH lamps. Drawing only 283W for 250W pulse starts they're not anywhere near as bad as the common 400W MH are but I think some good fixtures with the newer generation of 4 x T-8 lamps & ballasts will replace them giving me almost equivalent light at half the power consumption. Sadly T-5 HO fixtures don't fit into the equation unless I went to a 3 tube model and I still may if a good price can be found on what's considered to be a non-conventional # of lamps per fixture.

One would never want to see the utility bills on this building but I've sure reduced them with improvements over the past year. When it gets cool enough to work on that hot black membrane roof again I can get back to my evaporative roof cooling project in which controllers selectively mist cool water on the roof. Available water coming into the building has limited me to applying it only in two smaller select areas and that took almst 3/4 mile of 1" PVC pipe. On the controllers I designed for this I did actually get the PC Boards commercially made due to sheer component count (lots of analog with conventional CMOS logic) and the fact that if I can get the city to bring me in a few more 4" mains I'll be extending coverage to include the rest of the building. Even though there are many systems already on the market to do this my design includes some rather unique features thus, as with several other projects, I'm hoping to pull simple patent papers before I go sharing actual design.

Anyone could construct one of these and they're effective even in some residential situations but not anywhere near as much as they will be when your roof consists of balck rubber membrane.

 

Are you saying you spray water on to the roof when hot in order to transfer heat to the atmosphere, and so reduce internal heat ingress in to the building?

You don't have water baths for that do you - as per most evaporative coolers?

 

It's commonly known as evaporative roof cooling and has been around in one form or the other since the very early 1900s.

It works in two ways - the evaporation not only cools the hot roof surface to slow ingress but also takes some of the heat residing below it along for the ride.

In engineering numbers it takes 26,100 btu to turn 3 gallons of water into vapor, thus for every 3 gallons you can evaporate you're preventing &/or removing about 2 tons of heat from the transfer between outside and inside.

In the old days they'd actually build the roof such that it could hold a pool of water but this is far from as efficient as careful timing of just keeping the roof slightly wet. Common in ground lawn sprinkler components,valves and PVC pipe are used. You have to paint the top of the PVC pipe with some white latex paint as it isn't UV resistant, cheaper to do it that way than to buy the true UV resistant stuff.

Four LM335Z temp sensor ICs spaced about monitor the roof temperature and when it reaches or exceeds a set temperature level the system comes into efffect, making a cycle of misting the entire area then back to a waiting state if the roof temp has fallen below the setpoint or starts the cycle over again if it hasn't. Due to water pressure and plumbing loss limitations it has to be done in sub-zones such that no more than 4 or 5 sprinkler heads are on at any one time.

So far I only have two zones, each containing five sub-zones of four sprinkler heads each. Once a complete zone has cycled a handshake signal is sent to a duplicate controller box that handles the next zone. It does it's thing if it needs to then handshakes back to the first controller box.

 

I wasn't quite awake when I pade my last post, tha actual number is:

3 gallons evaporated = 26,100 B.T.U.'s removed (a bit over 2 tons)

 

Yes I guess the problems with the old pool technique would have been numerous - legionnares, algal growth, leak risk, maintenance access. Now I guess the biggest risk is a leak, which is probably best detected by a flow rate, or alleviated by having redundant shutoff valves - or maybe using float sensors in run-off pits.

A building roof can harbour a lot of tech from HVAC to gensets to batteries to antenna/comms to solar pv to wind gens.