This form does not yet contain any fields.

    Energy Management Systems

    Have you heard the phrase 'you can't manage what you can't measure?' This is true for your data center, and having an energy management system in place with your hands on the reigns gives you a good means to measure and manage. An energy management system (EMS) is to let you see where your energy is going as well as to help control and ultimately optimize the performance of the many seemingly disparate systems on your site.

    How is this different from the controls systems?  An EMS is more comprehensive as it looks into the energy use for the many systems and also using the utility energy use.  It can be broken down into certain sectors to help separate groups take more responsibility for their energy utilization.

    A good EMS will take you beyond kWh and natural gas to look at your fuel oil and water consumption, sewer discharge costs, and even to manage your vehicle fleet consumption. It can be used to do a number of great things for your data center, including: performance benchmarking; emissions reporting; site-wide power quality; public kiosks; and customized reporting of any of the information it receives.  That EMS might be so good that you’d like to expand it across your enterprise – and it would be able to.  A not-so-good EMS might do one or two things very well, but it may not be robust enough to cover the many facets of your business.

    The EMS is a software package and it has limitations. If you do not have the meters and reliable utilities to feed your system you cannot expect to receive accurate results, let alone trend future energy costs. Implementing an EMS should be as important as selecting the system to use. Finding the gaps in your energy measuring can be found and closed and discrepancies between your meters and utilities can be resolved as you work to bring the EMS online. The implementation process will let you see how to manage the EMS and how you expect to do reporting and future trending.

    The EMS can help you budget your energy needs year to year using month to month results. The energy forecast can be normalized for weather conditions and aid in giving justifiable reasons energy use.  While the software can do all of this and plenty more, it cannot mitigate the energy use – it can only point the way to efficiency.  It is still the responsibility of the user to understand the data, step in, and change operating methods and/or behavior.   



    Warmer Chilled Water on Existing Systems

    So we are looking to save energy, and one of those ways is increasing the chilled water temperature.  But how does that save anything?  Increasing the chilled water temperature brings it closer to the condenser water temperature and thereby reduces the amount of work (lift) a chiller has to perform.  Since the chiller usually has the largest energy demand in a chilled water plant, any reduction is almost a direct savings of energy.  And increasing the chilled water temperature is part of that plan. 

    But what are the implications?  As long as you are looking to keep the same temperature difference between the supply and return water, the amount flowing will stay the same, which means your pumps won't miss a beat.  Downstream, you will need to realize how your equipment will be affected.  The computer room air handlers (CRAH) units and other air handling will need to be reviewed to ensure that they can meet the cooling requirements, but also the dehumidification needed for the spaces served.  The CRAH units and air handlers are designed around a supply and return water temperature, and deviating from that may derate the unit, reducing its overall cooling capacity.  More importantly, air handlers aimed to dehumidify outside air for use inside will take less moisture out of the air than before.  Often this becomes the more important limiting factor for how high you can take that chilled water temperature.

    How should it be done?  Often this question should involve the managers of the chilled water equipment, as they may be able to tell you immediately if certain areas will likely see spikes in temperatures during peak performance days.  Stepping up by one or two degrees F for a week at a time while continuously monitoring the system provides the least risk.  As the chilled water supply creeps upward, outside air, building usage, and any other factors affecting the load need to be meticulously tracked.  Missing part of the load due to renovations, office swapping, or new added equipment can be detrimental to comfort. 

    Beyond the comfort level for the occupants, you will need to look at your process equipment limits: 

    • For water cooled equipment: if equipment has an intermediate heat exchanger, external or internal, the rise in temperature may not yield the best performance for the equipment.  Once the temperature starts to rise, the process equipment may not notice until it becomes 100% utilized - and at peak cooling demand.  If the equipment has to reduce the speed of its processes because it is too hot, then an increase on chilled water will exacerbate that problem.  However, it should be noted that most equipment will have a rating with limitations for water cooling. 
    • For air cooled equipment: ASHRAE has recommended and allowable operating limits, and in most cases this follows the Class 1 & 2 operating environments.  The CRAH units will need to be reviewed to see the impact to the air cooled equipment, but more importantly the data center rooms will need to be studied to ensure there are no hot areas that cannot meet the ASHRAE conditions for that data center. 

    Glycol & your system

    Using glycol for your chilled water system will have dramatic effects and understanding how painful it can be to your efficiency is a must.

    Why are we using glycol? In most applications the use of glycol is to lower the fluid (most often water) freezing point. The more glycol that is used, the lower the freeze point is depressed. This is great if you are aiming to produce ice or for a similar application.  However, for data centers this is not the likely goal. Glycol may be added to prevent freezing of piping that is exposed to low outdoor conditions, but this should be reviewed to see if it is necessary.  Often it can be added as a precautionary measure for piping exposed to low temperatures, but when the fluid is expected to keep moving there is usually little chance for it to freeze.  It is when the pumps stop and the fluid rests where the most danger of freezing occurs.  And even then insulation and heat tape are preventative measures that can keep the fluid temperature from freezing. 

    So perhaps you will reconsider a new chilled water system without glycol, but what about an existing system?  Most chillers are designed at ARI standards of 45 degree F supply, which is 100% water.  When the glycol amount reaches about 20% of the concentration, there are adjustments that need to be made for the chiller and the system.  The chiller evaporator can be designed to help counteract the losses from glycol by having a lower pressure drop and more surface area.  These adjustments also maintain a lower kW so that the compressor horsepower does not change.  For commissioning the lower pressure with glycol, the default settings will need to be adjusted to prevent nuisance shut downs of the compressor.  With glycol in the system will require at least annual testing to look for the pH and alkalinity to look for degradation of the fluids in the system.  Once the levels begin to change, the breakdown of the materials, such as piping, pump impellers, and heat exchanger surfaces, begins to accelerate. 

    Now you may want to know what the capacity loss might be of your chilled water system.  While there are small differences in glycol brands, as the concentration increases so do the losses.  This is because the heat transfer will decrease since glycol is less conductive than water.  But how much am I losing?  Here is a quick guide to help from a glycol manufacturer based on the two most commonly used glycols:

                                 20%  30%  40%  50%  60%

    Propylene    0.93   0.90   0.87   0.83   0.76

    Ethylene      0.90   0.86   0.81   0.76   0.71

    From this it is easy to see that using more glycol than necessary can be more costly to efficiency than most people realize.  If an existing plant decides to lower its propylene glycol concentration from 40% to 20%, the savings are about 6%; for a simple example this translates to:

    6% savings on the chiller operation for a 1,000 ton chiller at 0.45 kW/ton

    6% x 1,000 tons x 0.45 kW/ton x (24x7x365 hours/year) x $0.08/kWh = $18,900 per year. 

    That is a savings of 236 kW per year and $18.9 per ton per year based only on lowering the amount of glycol from 40% to 20%.  Keep in mind that if you eliminated the glycol completely, the amount of savings increases to 510 kW per year and almost $41 per ton per year. 

    Can we just take out the glycol on an existing system?   In every case the system will need to be examined to determine whether there are risks for freezing of the fluid.  If the glycol can be removed or decreased, there will be an increase in the amount of heat transfer capacity that the chilled water system will have.  The evaporator will operate more efficiently, increasing the overall heat transfer of the chiller, and the coils and other equipment adding heat to the system will be able to do so more efficiently as well. 

    Using glycol can be a safeguard against catastrophic freezing failures but those who pay those energy costs may want to explore ways to decrease or eliminate the glycol in their chilled water system. 


    ASHRAE 2012 Winter Meeting Notes

    I thought I’d provide a few highlights from what I learned and heard at the ASHRAE 2012 Winter meeting in Chicago.  

    ASHRAE has gone through some rebranding, including a new logo & tagline: Shaping Tomorrow’s Built Environment Today.   

    Of data center interest: Technical Committee 9.9 ( had two meetings at the conference; for more information it may be easiest to contact each voting member to ask questions:

    • Sunday, January 22 – Program, Handbook, & Research
      • N. Gangemi outlined some key dates for submissions to present papers and seminars for future meetings: 
        • ·         Feb 13 – Seminar submissions due for San Antonio (June 2012)
        • ·         April ~18 – Conference paper abstracts and full technical papers due for Winter Conference 2013
        • ·         June ~3 – Seminar submissions due for Winter Conference 2013
      • C. Kurkjian gave a short briefing on the handbook going to continuous annual updates for the online version.  Volunteers to help with the handbook are always appreciated. 
      • M. Patterson and the research subcommittee walked through a few items as they relate with working with other technical groups, including TC 4.10.  


    • Monday, January 23 – All other topics
      • J. Glass gave introductions and talked about how best to leverage the increasing membership of TC9.9. 
      • R. Pavlak introduced a working group from Standard 90.1; this group heard about the number of concerns between TC9.9 and the 90.1 mechanical working group subcommittee.  Several points were made to move things forward for future inclusion of TC9.9 regarding the data center spaces included in 90.1.  J. Glass and R. Pavlak would appreciate volunteers to help with inter-committee communication. 
      • R. Steinbrecher walked through a TC 9.9 white paper on revised thermal guidelines as part of the IT subcommittee report. 
      • D. Quirk discussed issues of NFPA with data centers. 
      • N. Gangemi relayed information and asked for future topics to be submitted, keeping in mind the dates needed for each of the future ASHRAE meetings (above). 
      • M. Patterson gave an overview on PUE and the step-by-step process to agree on a standard definition (Green Grid). 
      • J. Bean provided a brief trend of the data center industry. 
      • R. Johnson gave an update on the EU Code of Conduct and conveyed that in certain places the codes were rather lax. 
      • D. Beaty talked about the publications subcommittee progress – further details on new and revised books can be discussed with him. 
      • R. Schmidt gave the results and observations on contamination. 
      • M. Patterson gave a brief summary of the research discussions held the day prior. 
      • J. Rutt provided a short review of SPC-189 with a few items to be discussed as potential data center concerns at a later time. 

    Additionally it should be noted that the Standard 90.1 committee approved a measure to allow PUE to be used as a means of energy compliance.  This will likely be up for public review in a few months.  This will be an important piece to track, as it may be introduced in future versions of LEED and other energy standards.


    Seismic Ratings

    When do you need to have seismic rated equipment?  Should you have all equipment rated for an earthquake or just redundant equipment?  What equipment would be rated for a seismic event?  Of course the answer to these questions is - it depends. 

    The need for seismic rated equipment depends on the level of risk that can be accepted.  The first place to start would be to investigate the seismic zone at the location of the data center.  The applicable building codes will have guidance based on the zone, which will also specify the seismic requirements for life safety equipment.  This can be a good place to start to guage whether the equipment should be as robust to keep a facility, or just certain areas, operating after an event. 

    Looking at the seismic needs for only the mission critical areas is one way to cut the cost of providing seismic rated equipment versus typical equipment.  Another way to cut costs is to provide only seismic rated equipment on the equipment needed to provide the bare minimum of power and cooling.  This is a risk that must be recognized, as this would neglect occupied areas and ventilation. 

    Should your redundant equipment be seismic rated?  If it is not, it would not be considered truly redundant.  Perhaps that is a risk to be taken in a lower earthquake region, but otherwise this may not be the wisest recommendation. 

    While your equipment may have a seismic rating the other systems, such as piping, should be considered to have the proper supports.  After all, your equipment will not be able to operate long without the rest of the interconnecting systems to support it. 

    Although this doesn't have much to do with energy efficiency it has everything to do with maintaining operability after a seismic event.  When an earthquake happens, reliability still becomes the first goal of most data centers.  With seismic rated equipment checking and fixing becomes mcuh easier and the number of replacements is reduced, which is ultimately greener.