The Trend Towards White IT Racks: Should You Make the Move?18 min read

by | May 27, 2015 | Blog

Cabinet manufacturers report a growing percentage of their sales are for white server cabinets and we are starting to see some white equipment at conference exhibit halls, where just a couple years ago the prime variations tended to be flat black versus rippled black versus glossy black, with the occasional novelty color reflecting a marketing ploy to exploit a particularly high profile or high buzz end-user customer endorsement. So what’s behind this apparent trend and, if you haven’t already, should you take the plunge into the big white-out? The reasons for deploying white equipment in the data center have been regularly articulated by now: reduced lighting energy, better working conditions in the data center, and a more attractive appearance.

Aesthetic Considerations

Is a white data center really more aesthetically satisfying? I have heard both sides of this one just often enough now that I would have to say we’re getting into the area of personal preference where there is not a lot of room for swaying opinions. For every observation that white cabinets create a more harmonious visual experience in a space that tends to be mostly whites, off-whites, and very light grays, there will be a counter-observation that all that white makes the data center look too sterile. Furthermore, I would suggest that as far as aesthetics are concerned, unless you are running an enterprise disaster recovery data center out in the middle of nowhere, you’re opinion on aesthetics probably isn’t all that important. Rather, what appeals to your customers or board of directors is going to be the only opinion that matters and, unless they have actually articulated these preferences, you probably have a 50/50 shot of getting it right. The other reasons for considering white server cabinets, fortunately, are a little easier to articulate and define.

Potential Energy Savings

White server cabinets will contribute to a lower data center energy bill. This effect is a result of light reflectance value (LRV) differences between white and black surfaces. White will reflect somewhere around 80% of the light while black will reflect somewhere around 5% of the light. To give these differences a little perspective, consider the lighting requirements section of the new LEED for Data Centers scoring system.

6. If furniture is included in the scope of work, select furniture finishes meeting the following thresholds for area-weighted average surface reflectance: 45% for work surfaces, and 50% for movable partitions.

7. For 75% of the regularly occupied floor area, meet ratio of average wall surface illuminance (excluding fenestration) to average work plane (or surface, if defined) illuminance that does not exceed 1:10. Must also meet strategy E, strategy F, or demonstrate area-weighted surface reflectance of 60% for walls.

Clearly, black server cabinet LRV will not meet the requirements, while white server cabinet LRV exceeds the LEED requirements. Interestingly, in much of the LEED scoring rules, Division 27 products (server racks and all that associated network infrastructure hardware) are excluded. The lighting section is one of the few scoring opportunities for these products. Server cabinets, if they are part of the original architectural plan, are included in the furniture category and must meet the 45% or 50% LRV minimums. Airflow containment barriers, on the other hand, assuming they are bolted to the floor, fall into the wall surface category of item number 7 above and must meet a 60% minimum LRV. But if you’re not planning on applying for LEED certification, why should it matter? The thresholds in the LEED scale are just illustrative of a solid engineering correlation between surface color and lighting energy efficiency. This correlation is further substantiated by most anecdotal evidence of reports of 25-30% lighting energy savings in data centers with white cabinets versus black cabinets. If the lighting component represents 5-10% of the total data center energy budget, then a 30% lighting energy reduction equates to something like a 1.5 to 3% savings off the total data center energy usage. That savings value correlates nicely with the two points that LEED makes available for the lighting section of its scoring standard. Is that worth it? Given that there is likely no premium or very little premium for white powder-coated products, there is no economic reason not to make the switch to white cabinets and containment structures. While the ROI will not likely justify pulling out all the existing black products and replacing them with white products, for normal replacements, additions or new spaces, the ROI, while small, will always be positive and immediate or almost immediate.

Better Working Conditions

As for better working conditions, the anecdotal evidence here carries a little more weight: it’s my anecdote. While I did not spend every day of my career in the back of cabinets doing punch-downs and setting up and maintaining networks, I did spend enough time back there arranging temperature and pressure sensors and setting up data monitoring and collecting networks to affirm it’s a lot easier to work inside a white cabinet than it is a black cabinet. If we go back to the LEED scoring criteria for lighting, we see a maximum illuminance ratio between walls and work surface established as 1:10. Without getting into a lot of photometry minutia, suffice to say that illuminance (units can be either lux or lumens, or sometime phots, none of which is critically important here other than knowing the labels means you’ll know what is being measured if you ever see them) indicates how much of the incident light illuminates the referenced surface in a calculation that correlates to our human perception of brightness. What’s important here, however, is that the inside of the cabinet would be defined as the “work surface” and the standard tells us the surrounding walls of the room should not be more than 10X darker. Obviously, a black server cabinet doesn’t make a lot of sense in those terms. Furthermore, compare standard engineering guidelines for lighting for areas of easy office work or classrooms at 250 lux versus detailed mechanical work at 1000 lux and very detailed mechanical work (punch-downs, tracing cabling and port labels??) at 2000 lux, and we can see that it is recommended we give our workers a 4X to 8X perception of brightness inside the server cabinet as compared to in their cubicles. The numbers confirm what I have learned from my own experience of setting up thermal data gathering equipment in white cabinets versus setting up cabling and monitoring displays in the old days in black cabinets. How long do you think it takes to find a flashlight in a conference exhibit hall during the pre-show weekend hours or in a customer’s computer room after hours and the night security guy is using his flashlight four or five flights up in his search for full candy jars on admins’ desks?

Heat Energy Considerations

Finally, I recall reading somewhere that white cabinets had an advantage over black cabinets by virtue of the amount of light absorbed by black cabinets are converted to heat energy, thereby increasing the total heat load on the mechanical plant. I doubt that is going to be a very consequential impact because so little of the energy applied to a light emitter actually converts to light. For example, for incandescent light, for every 100 watts consumed by the light only 2 watts is converted to light and the rest is released as heat. For halogen lights, 100 watts of energy produces 3.5 watts of light and 96.5 watts of heat, and fluorescent bulbs will convert 100 watts to 8 watts of light and 92 watts of heat. Therefore, if white cabinets can reduce light energy by the 25-30% claimed by various data centers who have studied this (HP Wynyard, Cisco Richardson, TCS), then there is a relatively significant decrease in non-IT heat load directly from the lights, but not so much from the conversion of light absorbed by black steel or aluminum.

Conclusion

So should you make the move to white server cabinets? It probably doesn’t make much sense to convert an existing data center full of black cabinets to white cabinets. However, if you’re not one of those non-negotiable fans of the aesthetics of black cabinets, and the premium for white cabinets is somewhere between non-existent to negligible, then the data would suggest it makes good sense for new spaces or for incremental technology refreshes.

 

Ian Seaton

Ian Seaton

Data Center Consultant

Let's keep in touch!

2 Comments

  1. Anonymous

    Has there been a compare of white cabinets vs LED lighting from an energy cost savings perspective? I would think LED lighting would far exceed benefits from light reflection of white racks especially given how open IT racks are For airflow and cable egress. Also if you have white racks should you also have white IT equipment as well?

    Reply
  2. Dennis Kniery

    Another selling point for the better lighting when using white cabinets is the human error factor. Assuming we still believe that most data center issues are caused by human error, I submit that the people wiring server cabinets will make fewer wiring errors when the labels are easier to read. It would be interesting if anyone has either subjective or measured performance differences to support this assertion.

    Reply

Submit a Comment

Your email address will not be published.

Subscribe to our Blog

Archives

Cooling Capacity Factor (CCF) Reveals Data Center Savings

Learn the importance of calculating your computer room’s CCF by downloading our free Cooling Capacity Factor white paper.

Pin It on Pinterest