View Full Version : Trigeneration along with "conventional" refrigeration - hybrid cycle help!

25-02-2013, 12:54 PM
Hi all, first time poster, (relatively ;) long time reader.

I'm currently working at a University and we're involved in an energy efficiency project geared towards several Agrofood industries.

Long story short, I’m to develop a trigen system for a fish freezing industry. After analyzing their monthly electrical bills, I’ve realized that their “base-load” chiller, a Mycom VSD 200 HE (ammonia based) is responsible for roughly 40% of their yearly energy consumptions.

Initially I was hoping on going to an ammonia absorption machine that would enable me to completely meet his rated output, but after some calculations I had to abandon this idea due to the low COP –would force me to have a cogen prime-mover with an “off the scale” thermal output (the electric chiller with a reported COP of about 2 – given by the Mycom soft, consumes about 100 kWe giving 200 kW cooling; this would mean having a 400 thermal output to feed the absorption chiller…so many wastage is a big no-no these days ;)).

Now I’m looking at more “modest” prime-movers just to ease the thermal loads during peak hours.
My initial thoughts are: a trigen (LiBr-H2O based - more common and with better COP’s) system where the refrigerant – water would go in to a heat exchanger that on the water side acts like an evaporator and on the NH3 side as a condenser. I’m guessing that this would “ease” the thermal load on the evaporative cooling towers, thus lowering head pressure <-> compression ration.

I’ve inputed a 10ºC drop in the condenser temperature and indeed, the Mycom soft gives me a better COP (a 0.5 increase in COP).
But I also read that much care has to be made in lowering head pressures below certain limits, and that this is more effective on hotter months, have to put an extra pump to prevent flash gas, which in turn might provoke cavitation, etc and so on…. but what are your thoughts on this general concept? Is this doable?

Also, I would like to be more ambitious and make use of the remainder thermal output from the trigen to eventually superheat the NH3 (equivalent to raising the pressure) at the electrical chiller’s suction port. I was hoping that this would lead me to further savings as the compression ratio would be reduced, but if I increase the P or T in the suction, event though the COP goes up a bit, so does the electrical input to this chiller…this shouldn’t be happening as I’m lowering the comp. ratio so, what I’m I missing here?!

Lastly, are the values (COP, mass flows, etc) given by the Mycom software trustworthy? Because, not having NH3 flowmeters is a real nuisance in calculating COP’s….(can go with swept volume I guess, but how accurate is this method also?). For me to calculate the heat rejected to cooling towers, can I use the the liquid specific enthalpy of the ammonia on the low pressure side and the gas specific enthalpy at the high pressure side?
Like, m x (h_gas – h_liq) = rejected heat (heat from space + heat from compression) x 5 or 10% due to heat losses?!

Ps: many appologies for the big big post, but unfortunately I don’t know many people with proven hands-on experience that could advise me on possible technical barriers arising from this idea…

25-02-2013, 10:12 PM
During summer operation ammonia refrigeration plants with evaporative condensers operate at condensing temperature T cond.= Twet bulb+ wet bulb approach. Typically, wet bulb approach is 5-10C. You want to bilt additional refrigeration plant to cool condenser water. Am I right? If, yes. You can lower cond. temperature by 5c max. You can use this additional ref. plant just a few month per year. During cool weather ammonia plants operate at minimum allowable condensing pressure and additional ref. plant will be almost useless. What payback will be for this additional plant? 20, 50...100 years.

25-02-2013, 10:52 PM
Hi Segei, thanks for the reply. Well, not something that complex. I want to cool condenser water and to "pre-condense" or partialy condense the ammonia before going in the evap towers.
Altough the factory's situated in portugal, no quite canada, you do raise a good point: I have to accurately check average temperatures, more specifically wet bulb ones, which I suspect will be quite similar to the dry bulb as this factory is situated in between the sea and a river (few hundred meters from each..)

What about using part of the heat of the cogeneration to dehumidify (I think that RH is around 70% most of the year) air in the building's "envelope"? The high year round RH plus all the fish handling, floor washing etc putsthe factory at quite high levels of RH...if I could lower from +70's to 35 to 45, that would be a big aid on the electric bill (will check the building volume and report back with some quick calcs)


25-02-2013, 11:39 PM
I don't know what you mean by "precondense" or partially condense.
Wet bulb temperature similar to dry bulb if relative humidity close to 100%. You have humidity of 70%. So there is difference between dry and wet bulb temperatures. How are you going to get a big aid to the electrical bill by lowering humidity from 70% to 40%? Do you have free heat from cogeneration?
There are many ways to save energy in industrial refrigeration. However, optimization of the refrigeration plant operation can give us the shortest payback. Typically, 1-3 months. Any other approaches requires significant capital investments and payback can be as long as 100 years.:eek:

26-02-2013, 10:25 AM
What I mean by precondense or partially condense is to use the condenser/evaporator hx so that I can lower the compression ration on the electric chillers. due to the high price of electricity vs natural gas, I'm hoping that this would lead to some "cool" savings (wheter or not they are economically feasible that's a all other issue that I'll worry about in the following stage). This company is paying about >107 €/MWhe (excl. VAT) and the natural gas required to fuel the trigen would be around 40 - 45 €/MWh. Now, I've already simulated having the power generation (not COgeneration) during peak hours and just by producing electricity locally, this would save a few €/s. ofcourse I'm just interested in cogeneration because fully using the waste heat is the only way to save on primary energy (and greater capital savings)

This is why I'm looking for ways to size a proper trigeneration - if correctly done, trigen can bring big primary energy savings, which, for a country like Portugal that has to import all the energy it uses, is very important.

Now, as for lowering the RH in the building's envelope, I'm gonna go about to calc this today, but I'm hoping for good savings on the electric chillers - just think on the energy it has to be taken from the water vapor to turn it into ice

ps: I've been looking at their monthly electric consumptions and they do not present a clearly defined winter - summer patern with december and january being the only exceptions where they do not reach >150 MWh/month required elec. input)

26-02-2013, 02:57 PM
I'm focused on energy savings as well. However, any investment in energy savings should have reasonable payback. In North America this reasonable level is 2 years. If it is longer, companies will not invest. In your approach it will be much longer than 2 years payback.
To lower condensing pressure and save energy, you can install additional evaporative condensers. However, wet bulb approach should not be lower than 10F or 5.5C, otherwise additional condenser fans and pumps will use more energy than energy saved for compressors. Install VSDs for compressors(1 per suction temperature), condenser fans, evaporator fans. However, to maximize benefits of VSDs, right operating strategy should be chosen. This is practical approach. If you want to go with theoretical study, go with trigen.
P.S. Forgot to mention major energy savings measure which can give us up to 50% of total energy savings. This is lowering condensing pressure during winter operation. You don't need capital investments for that but you need knowledge and experience to implement this lowering.

26-02-2013, 04:07 PM
yeah, I'm kinda stuck between 2 different worlds - having to please the academia (my bosses) and presenting solutions that would be really appealing to the industry....unfortunately even with the high enough disparity between electricity and nat gas, the capital investments are always stupid high, as we have to import all the tech from you guys or from germany...:(

But I was really hoping that my study would turn out into something usefull, not just another project to go into the drawer...you see, I'm basing my values in the existing plant, but they said to me that this was any good, they could consider putting it from the star in another plant they are planning to build that will be very similar to the current one...so come on, don't discourage me so much :)

ps: on the VSD's they already have them in chillers that fluctuate their load a lot; not the case with the one I'm focusing on

26-02-2013, 04:32 PM
I don't want to discourage you. However, I found that optimized operation can give us a few months payback(because if limited investments) and VSDs can give us a few years payback(if you use them properly). Any significant changes in plant design(repiping, additional condensers and etc.) have long payback. You want not change existing plant, but put additional plant. This is going to be really long payback.
I suggest you to find the cost of additional trigen plant and compare with energy savings of 5-10%(this is the best you will be able to get) of the refrigeration plant. This will give you idea about real life payback.

mad fridgie
26-02-2013, 07:33 PM
Sorry not so much time to chat about this.
So just a couple of points.
Firstly you are making a mind set mistake, you have a solution, which you are trying to fit to the problem, instead of understanding the problem and finding a solution.
So a quick look at your solution, firstly you are placing your system in the wrong part of the cond cycle. Your Ct, is still determined by the wet bulb temp, regardless!
look at using you cond as desuperheates (large amount as high compression ration system), then look at using you plant to drop the Ct below wet bulb temp.
That should get you moving in the right direction.

27-02-2013, 12:35 AM
Actually, desuperehating is just 10% of total heat rejection. At lower head pressure it will less than 10%. Refrigeration load of new plant will increase by 20-30% due to additional heat from compressor work.

27-02-2013, 12:54 PM
hi guys, thx for posting.

Indeed I see that it would be hard to get "thermal users" to justify the trigen this way. I'm now back to square 1.
I've redone some calcs and I now get (ultra-rough) paybacks in the order of 6 to 8 years by substituting the base load chiller by an NH3 absorption one coupled to a nat. gas ICE. If this is to ever be set-up it would be in a new factory, so perhaps costs would lower even further

Just for you to no make fun of me and my long paybacks, here's the summary of an instalation that was recently built here in Portugal:

"In a Portuguese food company (xyz) will install a two-stage ammonia-water-absorption refrigeration system. The 1st refrigeration stage provides the customer with liquid ammonia and has a capacity of 900 kW at a temperature of -34°C.
The evaporator of the 2nd refrigeration stage cools down a brine of propyleneglycol/water from 0°C down to -5°C.
The driving medium is steam of 6 bar g.
Model: ARP
Refrigeration capacity LP: 900 kW at -34°C
Refrigeration capacity HP: 1.550 kW brine cooling
Brine inlet: 0°C
Brine outlet: -5°C
Driving energy: Steam
Steam pressure: 6 bar g
Condenser cooling: hybrid evaporative condenser
Absorber cooling: hybrid evaporative cooler
Start up: September 2009"

I'm actually trying to visit the instalations to get some "on-the-field" input....This is for a central food distribution warehouse that later ships for one of the biggest supermarket retailers spread across the country.

I'm bettin' that the payback for such an instalation wasn't 2 years....

The key issue here in Portugal is to correctly understand and simulate wether to sell, and at what time of day, or to "avoid" purchased electricity. this and ofrcourse the number of running hours of the trigen

The premium paid for selling electricity to the grid has recently been quite lowered, and many companies, such as this one are now only driving the trigen only during the day, recurring to the standard vapor compression chillers in nightime.

So, if these guys could do it, and I bet that they are not incurring in losses, I gotta pull it off also!

One request: give me ideas on what to do with the thermal wastage that goes from the absorption chiller? could I (cheappely) do something else useful with it?! (they don't need hot water lol)


07-03-2013, 05:31 PM
hi again,

so, noob question: if I allow my temperature lift to decrease, say Tcond = 35 and Tevap = -30 to a new Tevap = -15 (and with it, higher suction pressure -> lower compression ratio), why the hell does the electric consumption of chiller increases (even tough the cop increases too)?!?

On another entirely different subject, does anyone happen to know the electric input to the Grasso RC411 -- which I assume must have a few decades of age...?!

I read I think in this forum, that it was about 75 kW...I did confirm that with a "crazy" extrapolation regarding specific consumptions based on total volume of refrigerating chambers (at about same set-point temps) and the yearly consumption of their base-load chiller. coincidence or not, when I applied these values to the total volume of this different factory, it gave me 74.4 kW ;)

luck, or indeed we can take specific electric kWh /year.sqtf to extrapolate to different factories given similar climate and personnel "behaviour"?! ;)

07-03-2013, 05:58 PM
If you increase Tevap. compressor mass flow will increase, because refrigerant more dense at higher suction temperature. Increased mass flow means that cooling capacity increased. For example Capacity was 100 TR at Tevap.=-30 and it will be 130 TR at Tevap.=-15. Total energy use will increase, but energy use per TR will decrease.

07-03-2013, 06:13 PM
so the increased amps are due to the compressor having to "deal" with more moles of refrigerant right? offsetting this, these increased moles of refrig will positively impact refrigeration effect making the "balance" favourable to efficiency right?

cheers and thank you form the prompt explanation

18-03-2013, 05:00 PM
Hi all,

I've read that for each degree celsius that one lowers the condensing temperature, it can give gains of about 1 to 2 % to the compressor.... is this true?
If so, what would be preferrable, to lower the cooling water temperature or to lower the temperature of the of the refrigerant in a cascade heat exchanger (via an absorption cycle)?

Hoping to hear your thoughts on this,