The TCO for VoIP should include the electrical power bill. The cost to power the IP phones can be much higher when comparing one vendor’s power consumption vs. another vendor. This becomes apparent in the Tolly Enterprise’s report "Avaya 9600 Series Voice over IP Phones: Energy Consumption Evaluation versus Cisco Unified IP Phone 7900 Series".
The Tolly report came to three major conclusions:
1. In every scenario tested by Tolly, the Avaya 9600 phones consumed 40% to 60% less energy than the comparable Cisco offering.
2. Over 80% of the overall energy consumption of VoIP systems can be attributed to the VoIP phones.
3. Avaya VoIP solutions (gateways, servers, IP Phones) were validated by Tolly to use 45% less energy, produce 45% less CO2 and lower the electrical expense by 45% vs. Cisco.
So what does this mean in power and cooling? First, there are three classes of IP phone power consumption supported by PoE:
Class 1 devices consume fewer than 4 watts. There a very few IP phones in this category.
Class 2 devices consume 4 to 7 watts. A majority of the IP phones fall into this group.
Class 3 devices consume greater than 7 watts. An IP phone in this category usually supports 1Gbps, color screens and other advanced features.
A factor not included in the cost of power and cooling is the energy consumed and the heat produced by the power supplies in the LAN switch that is providing PoE. Nothing is 100% efficient. The LAN switch consumes 1 to 3 watts per port for PoE. The calculation we’re about to show does not include the inefficiency of the LAN power supplies, so that the actual savings by using a low wattage Class 2 IP phones would be even greater.
So let's create a situation for an enterprise with 1000 IP phones and create a cost comparison. The average electrical power cost in the U.S. is about 10 cents per kilowatt hour. The cost to cool the phones is an additional 9 cents per kilowatt hour (Kwh is 1 kilowatt = 1000 watts consumed in one hour).
Most enterprises would leave the IP phone on and powered for the full 24 hours and 365-1/4 days a year, for 8,766 hours per year. Consider a class 2 IP phone consuming 4 watts. It would require:
8,766 x 4/1000 = 35.1 Kwh/year
35.1 Kwh x 1,000 class 2 (4 watt) IP phones would require 35,100Kwh
35,100Kwh x 19 cents = $6,669 per year to power and cool 1,000 class 2 IP phones or $6.67per phone/year.
If a 7-watt class 2 IP phone is deployed, then the cost would be 61.4Kwh per year at a cost of 19 cents/Kwh for a bill of $11,666 per year. This is a cost increase of $4,997/year or $5 per phone/year for the 7 watt phone compared to the 4 watt phone. As energy prices increase, the advantage of the 4 watt phone only increases.
When evaluating IP phones:
* Make energy consumption part of the RFP document and product evaluation processes. A more expensive IP phone may have a lower TCO when energy consumption is included in the vendor evaluation.
* Determine the maximum and idle state power consumption.
* Select IP phones that turn off the display when not being actively used
* Buy single Ethernet port IP phones.
* Do not buy features that you may never use such 1-Gbps IP phones.
* Consider power management software in the LAN switch that turns of PoE ports during evening, weekend and holiday periods. Turning off the PoE during these periods can reduce the power consumption up to 60%.
In this time when going green is being hyped for the data center, especially the virtualization of servers, many in IT do not look at the small things like phones. Legacy phones usually consume less than 1 watt, 25% of the wattage consumed by the class 2 phone that is rated at 4 watts.
I have rarely encountered an RFP or system procurement negotiation includes the energy costs in comparing products for purchase. When I have presented this thought process, many consultants have added energy consumption to the IPT RFP.
I have written other blogs dealing with energy consumption and conservation. The most recent blogs are "Green ICT Questions, "Your 2010 IPT Energy Plan and "Green IT Initiatives: The Investment Value.