301_Hours's profile - activity

Your standby losses in the tank may be dominant assuming you have a storage tank.

Note storage tank losses are regulated under ASHRAE 90.1 in case that's applicable.

Also, the part load efficiency of the water heater may not be great if you're using the same size unit for the reduced flow.

For the larger plants they are designed to operate in parallel.

Like you stated the sizing becomes large if you want to avoid using the chiller entirely so parallel is needed.

Also, depending the size of the tank a chilled water reset temperature may be challenging. You'll have to monitor that to ensure you're not getting a bunch of unmet load hours.

Some additional practical items to consider that you may already know, but am noting for others are some sizing info, demand charge time frame and any redundancy you may be able to remove to reduce first cost.

You want to size the tank so the chillers (i'm assuming you're storing chilled water) don't run during peak hours. in very large systems (2+ million gallons [8M liter]) you may not be able to achieve this.

You need to size the tank for 10% additional capacity to account for themocline between the warm top layer and cold bottom layer of CHW.

Confirm your utility rate structure does in fact charge demand over a certain time period. I just finished an analysis for a 5M gallon CHW storage tank, which reduced the day time peak by 3,000 kW. However that 3,000 kW just came online later and i saved almost nothing on demand charges because the utility didn't care what time of day the peak demand occurs.

If you have multiple chillers, say 3 chillers sized for 500 tons with a building peak load of 1,000 tons. You may be able to remove the redundant chiller if your storage tank can produce 500 tons of cooling. That savings should help reduce your system cost.

Hope that helps.

I'm not an open studio user so I can't help on those specifics. As far as the control you would want to turn on the pump when the panel outlet temperature exceeds the return temperature off the coil or holding tank so you're adding heat to the system.

This isn't an open studio specific answer, but in other tools like TRNSYS I've noticed that if you don't turn off the pump when the sun is down the collectors will lose considerable heat. That should be mitigated somewhat in an evacuated tube system, but I'd check the flow rate. You could also do a temperature based control where the pump doesn't startup until the panel temperature exceeds a desired set point.

Also if the flow rate is too high you won't see higher temperatures as the heat is carried away too quickly likewise if your inlet temperature is too low.

Also, in northern climates solar hot water systems are often drained at night into holding tanks so that the working fluid doesn't freeze. That may help overall performance.

Finally, the angle of the collector makes a difference too. The best angle for overall solar gain is not necessarily the best angle for your system. Try more vertical angles to get more consistent year round performance. The summer will be worse, but the rest of the year will be better.

Hope that makes sense.

This is a big issue that hasn't been resolved yet as most laboratory applications are variable air volume. There was a 2012 SimBuild paper on this topic.

The work around is to come up with an aggregate schedule for the air flow based on the flow rate it should be. Simply put 12 hours at 6 air changes per hour and 12 hours at 4 air changes per hour would be 5 air changes per hour for 24 hours. Of course this doesn't capture peak demand charges correctly if that's a factor in the design and over sizes the water side because you're only providing 5 air changes rather than 6.

Note that the induction ratio is not linear with air flow and drops off dramatically below 50% of design flow.

As far as I know only IES-VE does this correctly today.

The technologies are technically mutually exclusive as there is no reason to have both as they generally perform the same function.

You don't have to have a chiller connected to a water side economizer. Just make sure to shut off the chilled water pump when the water side economizer isn't able to meet the load.

That said, if you're putting in a chilled water loop, a condenser loop, pumps, and a cooling tower, then you might as well buy a chiller and avoid buying the DX units. Unless of course you're using a dry cooler as your water side economizer. Even then you'll have a cooling coil with water and a cooling coil with DX. That will increase the pressure drop and consequently the fan power.

In reality it would be simpler to have a DX unit with an air-side economizer or a chiller with a water side economizer as it's the least amount of equipment to install and maintain.

Hope that makes sense.

Spandrel is opaque even though there is glass in the construction. The baseline for the spandrel area should be the same as any opaque wall assembly as required by your climate zone.

The inefficiency of the VFD is the only thing that should increase the pump power, which would be ~5%. If you have a load that generally requires constant volume, then a VFD would be worse performing. If it's a huge increase in energy then something else is going on in the model.

An improper schedule that wasn't active with a constant volume pump, but became active with the VFD might be a possible explanation.

Couple of other comments to add.

For ASHRAE 90.1 compliance it should be 1 cell per tower and 1 tower per chiller in a separate loop. Ie. don't put 2 separate towers on the same loop serving 2 separate chillers.

In reality towers can only have about 50% water flow in order to get proper distribution of cooling tower water over the tower fill. Please keep this in mind for condenser water pumps running on VFDs.

Running maximum cells vs. minimum cells makes a difference in fan power. One 10 Hp fan running uses more energy than two 10 Hp fans running at 50% due to the fan affinity laws.

If there is a passive chilled beam in EnergyPlus I would use that. The capacity is approximately the same, and it has a radiant and convective component. Although I'm guessing the radiant component in a sail is higher, and the convective component is lower.

You'd likely have to get an hourly report that shows the wall temperature, and then calculate the radiant heat transfer yourself.

I don't believe eQUEST does this. I don't know if eQUEST will even radiate on to itself if the geometry warrants it.

The only time I've seen this as a design issue is in urban areas where highly glazed buildings with reflective surfaces reflect solar radiation to another building increasing the solar gain of that building.

I would think that would be a more significant factor than long wave radiation from the surface of a building upon another building's skin.

So somehow 9 months has passed and I forgot to post what I submitted to GBCI, and which they have subsequently approved on this and one additional project.

Is there a way for me to upload a .pdf here? When I try to attach a file to this post it only allows images.

This link was sent to me today that may help answer this question.

Belimo has this tool for eQUEST and TRACE. Haven't tested it yet, but it's worth a try.

http://www.belimo.us/americas/sp_onli...

If anyone uses it, please post your experiences here.

you'll need to request a performance curve from the manufacturer, and then make a curve in eQUEST. It takes 33 selections to make the necessary curves for an air cooled plant so give your rep a heads up on that.

I don't think you need to worry about the condenser being separate as the heat rejection won't change substantially. Perhaps adding pump power as a process cooling load, and as a metered electric load would be the best way to approximate this. Maybe create an 8760 pump power schedule based on the hourly cooling load from the model.

You can create a water side economizer as a separate chiller on the same chilled water loop, and have a dry cooler selected as your cooling tower option in the condenser loop. Set the condenser pump power to zero.

To elaborate on this you'll need the following components in the chiller plant. Best to build it in the following order too:

CHW loop Condenser water loop CHW pump Condenser water pump Air cooled chiller Water side economizer chiller component heat exchanger as attachment to dry cooler heat exchanger pump as attachment to dry cooler Dry cooler Equipment controller Load managment

Then you'll want to create hourly reports for the pertinent chiller plant values to ensure that it's operating correctly.

Please reply to this post your experience here as this is not very straightforward, and 90.1-2010 is requiring more water side economizing. A reliable method of creating these plants should be recorded.

Good luck

This is not easy to do from the detailed mode, and requires you to manually edit the .inp file in a text editor.

Create a copy of the .inp file and then adjust the vertices of the geometry as required.

There is a lot that can go wrong with this process so take your time and work cautiously.

This is probably the single biggest disadvantage to using eQUEST.

Good luck.

I don't think this is possible as you can't have 2 systems serving 1 space in eQUEST.

I think you'd need to run the simulation twice, once with both fans combined, and once with just the smaller system. Then you'll need to piece the data together.

Or you could generate a custom fan curve that mimics the performance of the system, but this might cause problems in low flow conditions for the 6 months of the year where both fans are running.

Good luck with this challenging problem.

TRNSYS has a swimming pool component you can purchase from TRANSolar. I'm not sure if this will fully meet your needs, but you could build a TRNSYS model around the swimming pool component.

Hope that helps.

Does anyone have any idea about equipment power densities in Saudi Healthcare facilities?

I typically model 2 W/ft2 in a US patient room, but have no idea what it should be a in a Saudi facility? This facility is a lower quality community hospital so I'm thinking it will be lower than 2W/ft2, but I'm unsure how low to go.

I'd appreciate any insights or experience others might have.

All,

I'm starting a new model in Colorado with an elevation of 5,000 ft that has an air-cooled chiller plant. When running selections to determine the performance of the chiller I'm noticing that the performance of the chiller is worse due to the density of the air, which makes sense as more fan power is needed to reject the heat.

When I run a chiller only eQUEST model and change the elevation from 0 to 5,000ft, there is no change so I don't believe eQUEST accounts for air density at the plant level like it does on the system level.

So my question is, how do I derate my baseline DX system performance to account for the altitude difference? If I plug in 90.1 values into the baseline and actual selection values, I'm unfairly improving the baseline. I believe AHRI tests everything at sea level.

Thoughts?

Has anyone modeled the energy savings supposedly gained by pressure independent control valves?

Supposedly the temperature differential is better (larger) maintained, which would reduce pump energy and has an impact on chiller performance as well.

The only thing I can think of is to use a different temperature differential set point for the loop, but the difference would be a guess (2-3 degrees Fahrenheit)

Has anyone modeled HVAC system reaction times?

Specifically I'm wondering how fast HVAC systems (specifically active or passive chilled beams) will react to load and temperature set point changes.

Specifically I'm looking at a dorm room with passive chilled beams or fan coils.

LEED 2009 Core & Shell has occupant and equipment densities in Appendix 1 & 2 respectively citing numerous sources for the occupancy data.

I believe the equipment loads are building level, not space level as I have never seen a laboratory with a receptacle load of 1.4 W/SF. More realistic is 6-10 W/SF at the space level.

Thanks for everyone's help.

The exceptional calculation I put together for LEED was accepted, which added over 5% energy cost savings for my project.

The only independent supporting information I found was an ACEEE paper from 2005. Hopefully something newer will be published eventually.

Thanks again.

Autosizing can be problematic for any component in eQUEST, but it appears to be problematic randomly.

I've had to hard enter capacities for chillers, boilers, and preheat coils in the past to get them to work properly.

It's important to look at the hourly reports to confirm set point temperatures are being met for every hour.

If you're early in the design, you can use ASHRAE 90.1 guidance and over size your chiller by 15%, and your boiler by 25% to have a realistic capacity.

Good luck.

I'm most familiar with eQUEST for this issue.

Sometimes you have to hard enter coil capacities as the autosize routine doesn't work correctly sometimes. This happens frequently/randomly on preheat coils often. To meet the heating load eQUEST has the preheat coil, reheat coil, and possibly a baseboard in many of my models. I think it uses all of these together rather than individually as intended. In other words, the preheat coil may be set for 55F, but may not achieve that instead relying on the zone devices to meet the load. This doesn't always work so hard entering capacities can be a solution. I've had to do this with DX cooling coils too.

Changing reset schedules to gradually heat up or cool down after a nighttime setback can help too if the temperature difference is large.

Hope that helps.

Some other things to consider for modeling pools.

Public pools are often drained and refilled as frequently as every month. This is a huge energy and water consumption factor, often larger than the day-to-day use of heating and dehumidification.

ASHRAE Applications has a pretty good section on Nataoriums to help with load calculations.

Also, getting loads from RETScreen can work. TRNSYS has a pool component you can purchase from Transsolar.

There are DX heat recovery AHU's that can save a significant amount of energy, Dectron and Seresco are two common manufacturers. The savings from the heat recovery can be modeled(approximated) in eQUEST, but I haven't tried other tools.

Hope that helps.

Has anyone successfully modeled a regenerative elevator for LEED compliance?

I see a lot of literature claiming 50%-75% savings, but I'm not sure if/how I should change the schedule and/or peak load in the model.

Or should I just put the elevator on a separate electric meter, and take a percentage cut in an exceptional calc.

It's a 25 story building with 4 passenger elevators.

Thanks.