Reply to Thread

Originally Posted by mmassa

Saw your posts from 2012, Did you ever have any luck getting a more "Open Source" Plant optimization code?
We want to do it for a few clients, but not lock into Tridium or a vendors proprietary code.

carrier makes a chiller optimizer. works with all bacnet systems.

Saw your posts from 2012, Did you ever have any luck getting a more "Open Source" Plant optimization code?
We want to do it for a few clients, but not lock into Tridium or a vendors proprietary code.

Fisha - you have mail.

Lots of talk & words (it's all good) - for those who like pictures, see attached.

Originally Posted by CraziFuzzy

... writing cascaded controls that can tend towards a more optimized system.

...exactly

I am pretty sure they are not using Alerton controllers. They are an OEM Tridium and they deploy this system on a OEM Jace. At least that's what I have seen. I don't have any first hand experience with Climatech other than what I have heard which is that they are a huge successful company.

Originally Posted by fisha

Do you think these companies with their "black boxes" have automated what you are describing and are selling that software? Climatech is a well known well respected company and I find it hard to believe they would go to market with some voodoo that doesn't deliver. I could see a rogue mechanical engineer doing that though.

No, certainly not. They don't really have a 'black box' system, just code on standard Alerton controllers for the most part, it's just that they have a couple programmers that are good at writing cascaded controls that can tend towards a more optimized system. My personal experiences with Climatec are far from positive and have yet to see anything remotely 'advanced' come from them that actually delivered anything positive.

The level of data analysis I'm talking about is not something that's going to be easy to program in typical controller logic. It is something that needs to be written as a scientific application on a piece of general computing hardware.

Do you think these companies with their "black boxes" have automated what you are describing and are selling that software? Climatech is a well known well respected company and I find it hard to believe they would go to market with some voodoo that doesn't deliver. I could see a rogue mechanical engineer doing that though.

That is what I posted earlier. the best way to optimize a plant is to record the ins and outs of all the individual pieces of equipment, so you can predict what a given efficiency is going to be at a certain set of parameters, and choose the possible combination that give the best bottom line. It has to be a fuzzy logic type system that continually updates its knowns as they are experienced, interpolates between known data points, and continually evaluates conditions and determines which way to nudge current operating parameters. You almost have to have a full simulation of the existing plant running all the time, and try a few different combinations continually and if the result is better than current, move toward that.

I think you have to use field data.

My guess would be since you can almost treat required Pump Speed (and consequently energy consumed) as a known item (based on existing piping, and the equipment you're planning to run I guess you can make a rough calculation of required pump head). Comparing a total pump scheme using maximum tower cells (just flowing water, no fans) vs. full capacity towers (less pumping required because less towers to push water over, but significantly more cooling output) and then iterate for the in between to try and optimize... That's where I would start roughing in a skeleton for a program.

Interesting point to me is where do you start this whole thing? Obviously running max tower cells and no fans would be the least energy used on the Condenser Side (which it seems is close to how Hartman ends up executing), but there's the competing curve of the Chiller efficiency vs. Chiller Load vs. Condenser Water Supply Temp. Unless you've already been running this Chiller for XYZ number of days, or perhaps a full season, your only reference data for that calculation is Manufacturer Specs. I tend to use York for reference since I don't have any big chillers in my backyard to play with, but field data always seems superior to me... the end all question for you gurus, do you use mfr specs at all? or just develop a baseline of data and then start optimizing? or start with specs and slowly transition as data is accrued? The problem with starting using Mfr Specs is you're making (albeit realistic) assumptions about equipment operation, and you never get the full scope of how the Chiller(s) act at different Supply Temp & Load Variations because you're already trying to hone in on a certain area... potential for self fulfilling prophecy of the "best" CWST SP

It's almost like Chiller Optimization is a misnomer, it's Condenser Plant Optimization. Unless you want to add an entire other set of parameters to your system (i.e. the Chilled Water side of the problem), it almost seems to make more sense to treat that portion as a "constant" for any given point in time and optimize your Condenser Water to those conditions or perhaps a trend in those conditions.

So since reading this post I have been looking around at what companies are doing and offering. It seems that there are a few companies out there that sell and install a controller that has energy dashboards and high level algorithms that optimize plants. Climatech out of Arizona has one that they claim can lay over top of existing BAS systems. Kiltech Controls is another. It seems to me that there is a focus of variable speed condenser water pumps, varying chilled water speed based on valve position and not temperature or pressure, comparing the cost of fan power versus chilled water power and choosing the most efficient of the two. Most talk about primary flow systems and avoiding primary/secondary flow setups. I'd like to know exactly what these high level algorithms are but I guess they patent them and are trying to make money by selling there software.

Has anyone installed controls on a Hartman Loop system? Does it actually work as prescribed?

Originally Posted by tridiumtech

Another stab: since ice storage provides colder supply water one would think that the HVAC designers
could downsize the pumps and fans on a new design (lower 1st costs). But, boy, got to do your 'pencil
whippin' just right (max. ton-hr to store). Water: ~18 ft3/ton-hr. Ice: ~3.5 ft3/ton-hr



TES is really about decoupling the Load (demand) from the chiller (supply). At the CEM 'school' they tell
us that "TES is not energy savings". TES's key feature is efficiency. A conventional system typically has
its lowest efficiency at the time when it is needed the most. I would think that 'monster' USC TES system is a
bit of 'Load Leveling' (flatten the peak-run CH at more constant load) and 'Load Shifting' (shift peak hr AC
to off-peak hrs for better electrical rate). 'Crazy' ;-)

The efficiency gains are minimal, and would not likely ever justify the cost of a TES system. TES is all about the load shifting, due to not only the massive difference in price between peak and off-peak energy pricing, but also due to the better environmental conditions during the off-peak period. My point, in regards to THIS discussion, however, was that if you DO have the TES, it DOES make plant optimization that much simpler, because a lot of the "variables," are a lot less "variable."

Another interesting tidbit about TES, is that in most climates, the off-peak period is near dewpoint most the year, meaning cooling tower effectiveness is highly diminished, and air-condensing (either direct or indirect) becomes perfectly viable, again, simplifying the overall system, eliminating costly tower chemical treatments, etc.

Would be interesting to hear where the trade off is between ice and water is.

Another stab: since ice storage provides colder supply water one would think that the HVAC designers
could downsize the pumps and fans on a new design (lower 1st costs). But, boy, got to do your 'pencil
whippin' just right (max. ton-hr to store). Water: ~18 ft3/ton-hr. Ice: ~3.5 ft3/ton-hr

With the tank, you will only have to determine the peak performance point for the outside air conditions, and use that to control the load (by adjusting tank recharging rate) to operate the plant at that peak efficiency point.

TES is really about decoupling the Load (demand) from the chiller (supply). At the CEM 'school' they tell
us that "TES is not energy savings". TES's key feature is efficiency. A conventional system typically has
its lowest efficiency at the time when it is needed the most. I would think that 'monster' USC TES system is a
bit of 'Load Leveling' (flatten the peak-run CH at more constant load) and 'Load Shifting' (shift peak hr AC
to off-peak hrs for better electrical rate). 'Crazy' ;-)

Originally Posted by CraziFuzzy

Wow!

..get what yr saying now

Sure, I'll go with dev as a second bet

Space would be key. I think radiant floors (heat and cool) is another untapped resource. A radiant floor system I was part of opened my eyes to the advantages of storing energy right in the building floors to reduce chiller requirements. It operates as a "building source" heatpump. Recovering waste heat from the floors during the day to heat the building during the night. Cool concept and with ice storage allowed for some tiny chillers to heat and cool the building mostly off peak hours. Chiller pukes, they can run for a long time before any failure makes a real impact on the space.

Originally Posted by orion242

Would be interesting to hear where the trade off is between ice and water is.

Ice's ONLY advantage is space. The fact that the temperature has to be run so much lower makes the chiller itself much less efficient. If you don't have the SPACE for water storage, and Ice is your only option, then yes, it is better than NO load shifting. But if you have the space available, water will almost always provide better energy efficiency. The cost of building a prestressed concrete tank is actually not as high as some might think. If can also be done completely underground. As an example, USC has a 3.2 million gallon chilled water storage tank underneath their track and field stadium. USC's tank and system are obviously quite large, but the benefits are scalable to any chilled water system. on-peak/off-peak electrical pricing is a fact of life just about everywhere now, and thermal energy storage is really one of the best ways to reduce costs in that situation.

They only run their chillers at night. They run them at the most efficient power point for the given outside temperature/humidity. Once the tank is charged (the duration varies depending on most efficient plant capacity, and amount drawn out the previous day), they shut them down until off-peak time again the next night.

Originally Posted by CraziFuzzy

I was not describing a 'buffer tank'. I was describing a proper TES tank. We're talking about, depending on the size of the facility, million+ gallons of stratified storage.

Would be interesting to hear where the trade off is between ice and water is.

I was not describing a 'buffer tank'. I was describing a proper TES tank. We're talking about, depending on the size of the facility, million+ gallons of stratified storage.