With IT being more in the public eye than in much of its history, there is an imperative to do what we do as effectively as possible. This means getting the details right, and as IT professionals feel the pressure to operate at high efficiency, it creates an opportunity for manufacturers to market to them in a very persuasive manner.
From consumer gadgets through to enterprise hardware, the stated technical capabilities help a buyer determine whether a piece of equipment is a worthy investment. It is not uncommon to purchase a device such as a smartphone or MP3 player, only to find that the stated hard disc storage is out by 5%-10%, or the battery life only offers 10 hours of playback when running at unrealistic user settings. Sadly this same experience if often suffered by the enterprise buyer – whose need for accuracy and accountability is arguably far greater.
Efficiency Data – Just a Marketing Tool?
Energy management remains a hot topic in the data centre, and the various facets of powering mission critical IT systems demand intelligent management if they are to be optimised. The facts and figures associated with equipment purchasing therefore become particularly relevant.
IT professionals looking at UPS machines will undergo an accentuated version of the MP3 battery frustration. When reviewing UPS machines, a wide variety of figures are likely to be displayed in side-by-side comparisons.
Someone trying to calculate their ROI (return on investment) for a given UPS machine might find that there is as much as 3% difference between the ‘lowest’ and the ‘highest’ reported energy-efficiency data, in any given segment of the market.
End-users cannot benefit from this situation. While submissions for budget might rely on complex TCO (total cost of ownership) calculations to secure the relevant funding for upgrades or investment spent, the reality is that those calculations will assume a clawing back of cost over a given time frame. No-one wants to incur a conversation down the line as to why operational expenditure is higher than predicted. It comes down to unrealistic power consumption expectations.
The internal power-topology of competing UPS machines in a given market segment will often be very similar, and yet as we’ve identified above, the quoted performance can vary by up to 3%. This raises two questions:
1. How is it that this is possible?
2. Why is this permissible?
The latter question can be answered succinctly; the absence of any independent testing and certification authority results in OEMs having a relatively free remit to focus on ‘clever’ markets. What might ironically be called a ‘cold war’ emerges, as OEMs resort to one-upmanship around energy efficiency.
Sadly this is the situation that prevails, and there’s no immediate sign of it changing. And so, it becomes important to answer the first question, and bring some clarity to the situation.
Discovering the real numbers
How do data sheets vary so widely, whilst remaining perfectly legal to display? In essence, OEMs have three options when it comes to publishing their energy efficiency figures:
1. The guaranteed level
2. The nominal, base
3. The best possible
There are no prizes for guessing which option this brinksmanship leads towards for many firms. And there are a great many ways in which such figures can be manipulated. A non-exhaustive list of these might include:
- Supplying the nominal voltage, or in some cases even an input-voltage that is a little on the high-side. This contrasts sharply with the honest approach which delivers the guaranteed efficiency. To do this, the manufacturer must instead supply the UPS with the minimum voltage that the rectifier can accept. Clearly a conscious decision is made to seek the best looking data sheet.
- Disconnecting the battery, so as to eliminate any DC-bus losses. This provides a quite unrealistic comparison to the guaranteed mode of testing, in which the OEM would take into account DC-bus losses on minimum rectifier input voltage – a consequence of a connected battery.
- Maximising the efficiency on your data sheet can be aided by ensuring that the machine starts from cold, is tested quickly, and the cooling air is well below the nominal 35°C. The colder the air the better for getting a highly marketable score sheet. A more true-to-life test would see the machine reaching full operating temperature before the test begins, and be run with cooling air at a more typical 35°C.
- There are games to play in finalising the numbers. A manufacturer looking to ‘optimise’ its data sheet will add the maximum allowable tolerance, e.g. 10% of the total losses, as compared to the guaranteed number which will see the OEM deduct the maximum allowable tolerance, e.g. 10% of the total losses. Equally the best possible data-seeking firm will round-up their stats to the nearest half digit at one decimal place; the guaranteed approach will round-down the answer to the nearest half digit at one decimal place.
Making use of the Numbers
The long and short of this is that efficiency can be measured and interpreted in a variety of ways. The sad truth is that any IT professional seeking to purchase a UPS machine must approach the ‘facts and figures’ with due care and attention. It is a challenging situation, however, when supposedly factual data becomes such a subjective affair.
It is worth noting that the guaranteed data can provide an overly cautious set of numbers. One such example might be the fact that a UPS does not run all its life at minimum input voltage and maximum inlet temperature.
It is probably better, however, for a company to size the UPS cooling plant to cope with that disaster scenario, rather than finding itself poorly accounted for under stress conditions. However, most organisations would rather enjoy greater efficiency after purchasing its equipment than experience a deficit in their planned energy savings.
IT professionals purchasing new UPS equipment as part of a strategy for energy efficiency must scrutinise the data sheets they deal with. Inaccuracies on paper may have broader repercussions when it comes to analysing the buying decisions which were made on their watch.
And more generally, for IT professionals to move forward as part of a broader business conversation, it is critical that we have a strong and reliable platform upon which we can build.