Adding Mechanical Cooling to Indirect/Direct Evaporative Systems

On this website, I have discussed indirect evaporative cooling, direct evaporative cooling and systems that combine the two into indirect/direct evaporative cooling. As we saw from the last article, however, there will likely be many applications where additional cooling beyond what can be attained by evaporative methods is necessary to keep a space comfortable. In these cases, we need to add mechanical cooling into the mix. But it is important to understand how to do this—there are tricks that can preserve most of the energy benefit of the evaporative cooling systems you have designed into your system.

The first trick is to determine where to put the mechanical cooling coil. The temptation might be to think it would be a mistake to put it in as the last component in the air flow. This is because it seems you would then simply be removing a lot of the latent effect that you are putting into the air in the direct evap portion of the system. However, inspection of psychrometric processes indicates that this is not the case. In fact, anytime the dew point of the air leaving the indirect evap is lower than the desired design air drybulb, you will see an advantage in running the direct evap section if it is located upstream of the cooling coil. A quick example will demonstrate this effect: Let's look at a 12,000 cfm system providing 60º supply air at an extreme sensible ambient weather condition in Seattle, WA. We will first look at this system with a standard mixed air (20% OA) arrangement with a traditional cooling coil. Then we will superimpose a three-stage indirect-mechanical cooling-direct system and compare the energy performance:

(click for larger image)

In the above image, the traditional mixed-air psychrometric process is indicated in red and the evaporative process in purple (with blue indicating the mechanical cooling portion). Both systems start at the ambient OA condition of 95º/68º. The standard cooling system then mixes the OA with return air in a 20%/80% proportion and then cools sensibly to a 60º leaving air temperature. (The cooling process in the traditional system is shown as purely sensible, but in reality, the system would likely use a return air bypass configuration to allow the portion of the supply air to be super-cooled to achieve latent cooling and thus prevent any latent load in the space from steadily building humidity through multiple passes through this pyschrometric process.) Note that the sensible cooling load in this system requires about 20 tons of mechanical cooling.

In the three-stage evaporative system, 100% OA is first indirectly cooled to to the condition at point I/D evap + R2. Then, mechanical cooling takes over to point I/D evap + R3, after which the direct evap section evaporatively cools to a 60º LAT condition at I/D evap + R4. I have then shown a sensible heating process from the LAT to represent the zone load to demonstrate that this will provide a very comfortable resultant air condition in the space at I/D evap + R5. Note that this is true even if there is a significant latent component to this load (The resultant room temperature is approximately centered in the pink zone that represents the ASHRAE summer comfort envelope). For this analysis, the direct evaporative system is operating at full capacity, and the cooling coil is modulating to provide the desired LAT DB. (Since this system is 100% OA, we are not concerned about humidity levels building up in the space as in the recirculating system.)

The first thing that should just jump out of this is that the evaporative system requires less than HALF the mechanical cooling of the traditional system--while providing the increased ventilation benefit of 100% OA! And this is neglecting the additional latent cooling load that would probably be needed to maintain humidity levels in the space with the traditional recirculating system. To add to the IEQ benefit, the direct evap section works as an air washer and effectively increases the filtration of the air to improve IEQ beyond that of a traditional 100% OA system.

Now lets compare the three-stage system we just examined (with a indirect/cooling coil/direct arrangement) to that of a three-stage system with an indirect/direct/cooling coil configuration:

(click for larger image)

Two things should be obvious in this example. First, the mechanical cooling load is even lower than in the previous example: down to under 5 tons! Thats about a quarter of the load for the traditional system, and a little more than half of the load of the evaporative system with the cooling coil before the evap section. Second, and this is the secret behind the reduction in mechanical load, the air leaving this system is significantly closer to saturation than the previous example. In other words, despite the fact the leaving dry bulb temperatures are the same in both cases, in the latter case there is more latent heat in the supply air. The evaporative cooling process before the coil allows the system air to hold more latent energy but yield the same sensible condition for conditioning the space. A quick inspection of the comfort zone indicates this air is perfectly suitable to provide an acceptable comfort condition, even with a reasonable latent load.

Earlier, I said there should be an advantage to running the direct evap upstream of the cooling coil if the dew point of the air entering the direct evap was colder than the design air DB. In this case the dew point of the entering air is 53 degrees--which is quite a bit cooler than the 60 degree design point we are looking for, so thus we gain the advantage seen. What about the case where the entering air is too moist? Let's look at a system where the entering air dew point is well above the supply air temperature:

(click for larger image)

In this case, the OA enters with a 62º dew point. It cools through an indirect section to about 72º/65º, and then directly to a cooling coil to reach the leaving air temperature of 60ºF db. Since the enthalpy and WB lines are nearly parallel, it seems there is very little advantage to using direct evaporation to get the air leaving the indirect evap section to saturation and then cooling it. But, importantly, there is certainly no disadvantage, (other than the electrical draw of the pump). Also notice that again, this system provides a significant load reduction compared to a standard system even while providing 100% OA!

Running through that process actually shows a slight advantage for using the direct evap section:

(click for larger image)

This cooling advantage should be confirmed for your specific system since it is highly dependent on the latent capacity of the cooling coil, and is offset by the pump energy and some small increase in system static pressure when the direct evap media is wet.

However, keeping the direct evap pump running even in these conditions provides several advantages besides energy savings:

• Simplifies the control scheme
• Provides IEQ benefit of air washing
• Increases the life of the direct evap media by reducing cycling of the evap pump

Whether or not it makes sense to use the direct evap portion of your system in times of high ambient moisture is a decision that can change depending on the particulars on any given project. But if there is a net energy penalty for using this system when the OA dew point is high, one can see from this analysis that the penalty is slight and that it would only occur for very few hours a year.

Introduction to Indirect-Direct Evaporative Cooling

Now that we have covered the basics of the indirect and direct evaporative cooling processes, it's time to consider one more wrinkle--putting them together.

In the IDEC cooling discussion, I made the point that not only do we get a reduction in dry-bulb temperature as our airflow passes through the IDEC unit, but we get a reduction in wet-bulb temperature, also. And since we now have seen that the direct evaporative cooling process depends critically on the wet-bulb temperature of the air it is cooling, it seems we should get some advantage by running the air through the IDEC section, and then running it through the direct section. And we do:

(click for larger image)

As you can see, the resultant leaving dry bulb is on the order of 64º, which is better than the resultant of 72º from the indirect section alone, or 69º for the direct evaporative section alone. Now 64º degrees may not seem cool enough for typical cooling applications--and for most projects it probably isn't (although it is important to not that ASHRAE comfort conditions can be met with this leaving air condition in a predominantly sensible load application given enough air). But keep in mind that this is the performance on a design day. How many hours a year would you be able to meet the traditional supply air temperature of 55º? Lets look at psychrometric chart with Seattle bin data loaded into it:

(click for larger image)

A quick note of explanation: The vertical line at 55º is the economizer line--any climactic conditions to the left of that line can be used to create cooling air directly using OA alone (or mixing OA with RA) and thus require no additional cooling at all. The blue diagonal line along the 53º wet-bulb line is a conservative mapping of the direct evaporative regime. At any bin hours under this line, direct evaporative cooling can be applied to the ambient OA to achieve cooling air directly. And lastly, the red diagonal line above that is the indirect-direct evaporative cooling regime, where the application of both cooling techniques will provide acceptable supply air conditions (assuming about 70% effectiveness on the IDEC). And above that line, the indirect evaporative system can still be applied to greatly reduce the load on any supplemental mechanical cooling system, if used to meet the same 55º leaving air condition.

Two things jump out of this analysis: First, the vast majority of the hours are satisfied without using mechanical cooling. In fact, in Seattle, most hours are met with simple economizers--which explains the emphasis in our local codes on this cooling technology. You can even think of evaporative cooling as simply an enhancement to the standard economizer. The second takeaway is that there are still quite a few hours that are not met. How can we address this?

Well, one way is to play around with the leaving air temperature. If we supply some more air to the zone, we can provide warmer cooling air. Let's look at that same chart, only this time lets use a supply air temperature of 60º:

(click for larger image)

By simply providing for a little more air to the zone, we meet a much higher percentage of the bin hours; so much so you that can now consider a system without mechanical cooling, as long as the occupants are willing to accept a few more hours outside of standard comfort conditions a year. Granted, this additional comfort comes at an energy cost--the cost of moving that additional quantity of air. This cost is, of course, offset by the avoidance of mechanical cooling. But, additionally, we know from the previous chart that this additional air is not needed all of the time. A variable speed control on the fans would naturally bring the air volumes down during periods where colder air is achievable.

One of the things that should be obvious is that this analysis is greatly dependent on the local climate and elevation of the project. To evaluate how effective this cooling method is, you need to create similar plots for each project locale. And where you are in the state has a great effect on how well you do. For example, a cool-wet climate like that on the Olympic Peninsula sounds like it might be a good candidate. So let's see how it compares to Seattle:

(click for larger image)

It looks pretty similar to Seattle, as we might guess. How about a hot, dry climate like Spokane?:

(click for larger image)

That's a real winner! There's only a small fringe of hours outside of the range where indirect/direct evap works alone. So if Spokane works, surely Yakima must also be a great candidate:

(click for larger image)

Hmmm... There's quite a few hours outside of the indirect/direct evap zone. Good thing we did this analysis before committing to a evaporative-only system!

Indirect-direct evaporative cooling, either as the main cooling technology or as an enhancement to the economizer cycle is a technology that has wide application in the Pacific Northwest, even in rainy Seattle. But it is a technology that requires careful analysis--it's not as simple as throwing compressor tons at a cooling problem. With today's emphasis on energy efficiency and sustainability, it is a technology that deserves a second look.

There is certainly more to talk about on the subject. Future topics will include integrating compressorized cooling with an evaporative system, indoor comfort conditions, water treatment and maintenance, control of evaporative systems. and the role of return air in these systems.

Resources you may find useful:

Energy Labs Indirect/Direct System Performance Calculator (Simply the direct and indirect calculators linked together
Energy Labs Direct/Indirect Evaporative Systems Engineering Guide (booklet format)

Introduction to Direct Evaporative Cooling

Now that we have discussed Indirect evaporative cooling, let's move on to the next question: What is direct evaporative cooling?

Direct Evaporative Cooling is a process where air is sensibly cooled by the effect of the evaporation of water directly into the delivered air stream. This is typically accomplished by use of a wetted absorbent media in the air stream, most commonly Munters CelDek (pdf) or GlasDek (pdf).

Evaporative media in an Energy Labs Unit

This process has the advantage over IDEC systems in being much more efficient (with efficiencies in the range of 90% easily attainable), but with one major difference: Direct Evaporative cooling is an adiabatic process. This means that there is no energy added to or removed from the airstream. The enthalpy of the air is unchanged, even as the sensible temperature is cooled.

How is this possible? Well, essentially you trade sensible heat for latent heat. As you reduce the dry-bulb temperature of the air, you concurrently increase the humidity ratio of the air. What you lose in sensible heat, you make up in the heat embodied in the evaporative phase-change of the water.

What does this look like on a psychrometric chart? Take that Seattle design day* of 85º/67º db/wb. Let's bring in 22,000 CFM of 100% OA. If we select a direct evaporative system with an 89% efficiency, the leaving air temperature will be about 69º/67º. Note that the wet bulb is essentially unchanged. The chart of this process is below:

(click for larger image)

Essentially, for this service, you get about 32 tons of sensible cooling but and zero tons of total cooling. If you were to allow this supply air to warm sensibly to a room temperature of 75º, you would find that the space RH would be close to 70%, which would probably not be acceptable for standard comfort cooling applications. However, in high-sensible cooling applications, like, say, data centers, this method of cooling has great application.

Additionally, since the resultant indoor conditions depend greatly on the outdoor air conditions, Direct evaporative cooling can provide acceptable air conditions for much of the year in a cool, dry climate like Seattle. In fact, any time the ambient wet bulb temperature is 53º or less, the direct evaporative cooling can provide supply air almost identical to that off of a 55º cooling coil, with pressure drops at the media on the order of half that of a standard cooling coil! Direct evaporative cooling can used to essentially greatly extend the hours of economizer performance available on almost any cooling system.

But the benefits do not end there--because pre-cooling with direct evap systems upstream of a cooling coil can significantly decrease energy costs for sufficiently dry ambient conditions:



Resources you may find useful:
Energy Labs direct evaporative performance calculator
Energy Labs Direct/Indirect Evaporative Systems Engineering Guide (booklet format)


*Note: When applying evaporative systems, often it is necessary to consider the performance of the system at the ASHRAE evaporative design day conditions, in addition to the sensible design day conditions that we commonly use. And, additionally, it can use what would normally be unwanted space heat in the return air to provide beneficial humidification in times of low humidity.

Best Practices for Data Center Design

A 2006 study published by Lawrence Berkley National Laboratory, Environmental Energies Technologies Division examines best practices for energy-efficient data center design. Some of the best practices highlighted include well-understood and accepted practices like ‘right-sizing’ central plants and using hot and cold aisles. However, the article makes a strong case for the use of air-side economizers for the minimization of energy-using refrigeration and direct evaporative cooling for humidification. Read more here:

Best Practices for Data Centers: Lessons Learned from Benchmarking 22 Data Centers