• Google ..

  • Google Content

  • RE_Calendar

    November   2017
    Su Mo Tu We Th Fr Sa
    1 2 3 4
    5 6 7 8 9 10 11
    12 13 14 15 16 17 18
    19 20 21 22 23 24 25
    26 27 28 29 30
  • Air and water side economisers in Data Centers

    There is a common myth that recently pervaded the IT sector, namely the use of so called “Air side Economisers” or ASE later in the document. ASE consists in fresh air intake into the Data or Telecommunication Centres (Datacom as referred later on), in order to obtain a pre-cooling of the air. The recent publication of the technical specification of a big social network Datacom (available at http://www.opencompute.org/) further introduced elements of confusion.
    Nonetheless at least one viable alternative is available, namely the so called “Water side Economising” or WSE later in the document, that is a heat exchange process, in which the heat transfer media is water.
    I will try to discuss the pros and cons of both methods.




    ASE
    Multiple requirements are facing:
    - Need to reduce the energy costs and carbon footprint for cooling the Datacom, by exploiting cool outdoor conditions
    - Pressurisation of Datacom to avoid ingress of dust and other volatile organic and inorganic compounds (VOCs or VICs)
    - Possible heat recovery for heating adjacent living spaces.

    Installing air handlers (AHU) with multiple configurations of coils, and heat exchangers possibly satisfies two of three above requirements, while being quite energy hungry.
    A different approach would be by directly blowing outdoor air into the Datacom, after a filtration and an evaporative cooling treatment. The energy efficiency of the latter solution is undisputed. It is also true that many US Datacom centres are progressively reducing the classic compressor cooling and favouring the evaporative cooling (EC), but it is my humble opinion that EC still shows signs of immaturity, while carrying an intrinsic risk towards the integrity of the electronic equipment.
    Specifically, even whenever a good grade of filtration and pre-emptive maintenance are ensured, the presence of particulate (PM10 and PM2.5) may damage or create micro short circuits over the PCBs.
    In the greater areas of metropolitan cities, specifically in SE-Asia, but also in Europe at or around chemicals and petrochemicals compound plants, the presence of elements and VOCs was noted. It is ascertained that Chlorine, SOx, and Hydrogen Sulphide are causing creep corrosion, via corrosion, and formation of Sulphur crystals that may originate short circuits.
    Secondly, it must be noted that EC is consuming enormous quantities of water, which should avoided where historically water is scarce, and also for ethical reasons.
    Thirdly, it can’t be ignored that the latent heat added to the Datacom by the EC process technically makes controlling humidity impossible or extremely costly.
    Even where EC is not necessary, or “Air side Free Cooling” (ASFC) is used (that is plain admission of cold winter air), the pollution originated by the presence of chemical compounds in outside air, and the associated corrosion phenomena, are still critical points that can’t be ignored.
    Last but not least, the Latitude of Datacom is fundamental when ASFC is planned, as the ratio between implementation costs and energy saving shall be carefully analysed from a financial point of view. In fact, statistics of savings based on historical and climatological data do not guarantee similar or comparable financial results in the future.

    WSE
    The requirements are the same than ASE, but in this case the heat vector fluid is water.
    There are two universally accepted methods:
    - Outdoor air to water chiller, with Free Cooling exchange section. In this case, the on board microprocessor automatically switches between compressor and free cooling based on the actual weather condition. The process is total transparent to Datacom close control units (CCUs)
    - Water-cooled CCUs, connected to outdoors dry coolers, and built-in Free cooling coil. In this case the switching between compressor and free cooling is done via the CCU on-board microprocessor, by reading the return water/brine temperature from the dry cooler.

    The main advantages of this approach are:
    - Totally segregated Datacom volume. No possibility of VOC/VIC pollution ingress.
    - Easier installation, lower first costs, and quicker set-up compared to ASE.
    - Close circuit ensures no water consumption at the expense of a minimal cost of pumping.

    Disadvantages can be summarised as follows:
    - Again, historical and climatological data do not guarantee successful financial breakeven in the expected term.
    - The use of a secondary fluid (water or brine) reduces the overall efficiency, due to an additional thermal exchange (air to water to air).

    Mandelli Alessandro
    Comments 11 Comments
    1. NoNickName's Avatar
      NoNickName -
      Lauren, thanks for your input.
    1. malikansar19871's Avatar
      malikansar19871 -
      thnks for such a gud detail
    1. spoon man's Avatar
      spoon man -
      Looking at above info looks very true, have never worked on a system like this. Surely it is only cost effective on a low ambient temp period country.My company only works on chiller systems ev and free cooling (which is only cost saving if you have enough winter period).We work in countries such as dubai etc which obviously are hot. Look at our patented design for data centres at aquacooling.co.uk
    1. olgadrake's Avatar
      olgadrake -
      i can agree more with spoom man.
    1. pkagarwal's Avatar
      pkagarwal -
      i am pramod kumar agarwal. i am newly joined this group .please help me
    1. phamvuong146's Avatar
      phamvuong146 -
      thank you for your asei hope receive one more your`s
    1. mark957's Avatar
      mark957 -
      Quote Originally Posted by phamvuong146 View Post
      thank you for your asei hope receive one more your`s

      beautiful girl
    1. John@NRC's Avatar
      John@NRC -
      NoNickName, thank you for a very well written thought provoking treatment of the economizing possibilities of data centers. Per square foot data centers consume more kW's than any other space in a normal structure thus worthy of our consideration. I would like to interject a few thought into the discussion, if you don't mind. Over the course of my career data centers have migrated from ultra clean 60-65F environs to what the IT people(in my humble opinion joking) call shirt sleeve environments. Although I understand and accept the theoretical rational for air to air exchanges in practical terms it is almost anecdotal in terms of quantitative analysis, it leaves me with a warm fuzzy feeling but very hard to prove. Free cooling in water source exchanges such as evaporative cooling thru a cooling tower using condenser water as the primary refrigerant and a plate and frame heat exchanger using water as the secondary refrigerant has proven quantifiable cost savings. In data centers that my teams have managed in the mid Atlantic region of the US we have reduced operating cost by 40-50%/sq.ft. in terms of electrical cost and have gained the hard to calculate savings associated with an increase in equipment life. Always keeping and eye open for new innovation and having practical experience in it usage of the emerging concept of rack or vertical cooling, I feel it appears to offer the greatest return on investment that I have seen so far. By enclosing the IT equipment in towers the space needed to be tightly controlled is minimized and removes several of what I will refer to as parasitic heat gains i.e; lighting, structure, human, ... and allows a more exact control of equipment environs. The saving can be easily quantified which as Engineers we should all hold as the Holy Grail. Before I attempted to install any new economizers I would 'try' to preventing the IT type from viewing raised floors as raceways that they can fill up so completely as to prevent the proper movement of air, which I have found to be is the prime cause of system issues leading to increased cost of operations control issues. Again thank you for the thought provoking post.
    1. NoNickName's Avatar
      NoNickName -
      Thanks for your post. I would add few more cents to the discussion, by inviting you having a look at the most recent innovation at www.logica-system.eu .
    1. Shailesh Sonone's Avatar
      Shailesh Sonone -
      Shailesh hereRecently joined as application engineer , so lot of dought , confusion -- expecting guidense from u
    1. MANAR's Avatar
      MANAR -
      Thanks for this information
  • Recent Forum Posts

    Dj barman

    Whrilpool k20 overfilling

    Hi All hope i have posted in the right place
    As a bit a a dab hand at fixing anything i have been left responsible for a K20 ice machine.

    Dj barman 19-11-2017, 02:44 PM Go to last post
    Lambo

    Re: Ammonia two stage intercooler design

    Thank for effort but I already check that

    Lambo 19-11-2017, 12:41 AM Go to last post
    Brian_UK

    Re: Ammonia two stage intercooler design

    Not my subject but here are some items that may be of interest...

    http://bit.ly/2ivfrAh
    http://bit.ly/2B0US6G
    http://bit.ly/2zTbg94

    Brian_UK 19-11-2017, 12:01 AM Go to last post
    Lambo

    Ammonia two stage intercooler design

    Hi everybody,

    Where can I find explained in detail calculation principle for separator/intercooler for ammonia two stage plant? I am interested

    Lambo 18-11-2017, 10:58 PM Go to last post
    josef

    Re: trouble on my Cellmatic board

    Check Ground.

    Josef.

    josef 18-11-2017, 10:02 AM Go to last post