Maintaining cooling optimization for long-term energy savings: Everything you need to know
During our latest webinar – Data Center Cooling Optimization: How to maintain energy savings long-term, we reviewed how our cooling optimization program (COP) significantly reduces cooling energy consumption and demonstrated how to maintain energy savings long-term. We were joined by our partners at EkkoSense, who introduced their platform to a Canadian audience.
As no two sites are the same, our cooling optimization program takes into account the intricacies of each unique space as we develop and implement a solution to improve cooling energy efficiency. Through a set of customized energy conservation measures, we reduce cooling energy consumption, allow for increased IT load, and extend the service life of existing cooling units. Achieving substantial energy savings and maintaining optimal thermal conditions for IT equipment is the objective of our COP, and ensuring the continued efficiency and savings is the final component of the program.
Our adoption of EkkoSoft Critical is a natural addition to our COP as it allows us to maximize the energy savings that we achieve for clients, long-term. Thermal monitoring provides the real-time data needed to ensure the optimized state of a data center we’ve implemented energy efficiency improvements in is maintained, maximizing energy savings.
According to our webinar viewers, their top data center concerns included cooling capacity and level of cooling utilization, both of which are addressed in our COP and the Ekkosoft Critical platform. In this post, we’ll be sharing answers to some of our viewer’s questions on both our COP and EkkoSoft Critical, to give more insight into how you can successfully maintain long-term energy savings in your critical facility.
We looked at COP performance in a variety of data centers, and received the following questions:
What are the most common energy conservation measures (ECMs) that you apply to data centers?
In a recent webinar – 5 energy conservation measures to optimize your data center cooling, we looked at our most common energy conservation measures that we apply to improve cooling energy efficiency and reduce operating costs. The decision of which ECMs are applied is driven by the conditions of the data center and the input we receive from personnel responsible for the data center. Air flow optimization is one ECM that is always applied as it is key to achieving a thermally acceptable data center and maximizing energy efficiency. In a case study we shared during the webinar, we explain how the application of appropriate ECMs drove the electricity costs down by 38%, improved the PUE from 1.48 to 1.24 and saved the client nearly $39,000 per year on electricity costs.
Can you comment on the difference in the performance between cold aisle and hot aisle containment? What is your preference?
There is no right or wrong answer to this question. In many cases, the geometry of the data center dictates what can or cannot be done. For example, if the ceilings are 5 metres high it would be difficult to do hot aisle containment. In a slab floor site with ducted cooling, end of row doors on the cold aisle can work as well as full cold aisle containment. In legacy type spaces, there are all sorts of considerations to take into account. We would usually target cold aisle containment to maximize efficiency of the cold air delivery from the cooling unit all the way to the intake of the IT gear. However, fire protection/suppression, employee work environments and some physical layout issues do come into play when selecting what we choose. The budget, of course, always needs to be considered.
In new sites hot aisle containment tends to be more favoured however as noted above work conditions need to be taken into account. The bottom line is, any method that separates the cool and warm air is a plus and to try to quantify the difference in effectiveness would be difficult.
What PUE target would you like to see?
In our world, we look at cooling optimization and cooling efficiency as key KPI’s, not a target PUE as that is a value that can be easily manipulated. We do collect the necessary data to provide a partial PUE (pPUE) value, which includes IT load and cooling power draw. This is a relative value intended to establish a baseline and end of project comparison. In the timeframe we are doing our cooling optimization work the IT load typically doesn’t change, (and if it does that is taken into account) so any reduction in cooling power draw will show an improvement in the pPUE.
In an earlier webinar – Reducing Energy Consumption in The Data Center: Case Studies in the Real World, we reviewed case studies and referenced baseline and completed project pPUE. In one case, the pPUE was reduced from 1.71 to 1.41 and in another, the pPUE was reduced from 1.45 to 1.31. The more important aspect was the release of cooling capacity and power draw which could then be used for additional IT load, with no increase in cooling systems required, while reducing cooling energy costs by over $32,600 per year representing a 30% decrease
The experts at EkkoSense took our viewers through a walk-through of the EKkoSoft Critical platform; the following questions were asked:
How intrusive is the implementation and hardware requirements?
The platform has been specifically designed to be simple and non-intrusive for the user. Wireless sensors simply tie wrap/stick to the front of the racks, the wireless receivers mount to wall/pillar/cable trays and the EkkoAir sensors for the cooling units are installed in under 20 mins.
Does it integrate with other systems and devices?
Yes, we routinely interface with 3rd party devices via Modbus and SNMP and we are also able to communicate via oBIX. The capability to support 3rd party devices means Ekkosoft Critical can monitor generators, UPS equipment or other active equipment in the data center.
Is the live view of the DC floor a sort of DCIM? integration of Monitoring, Layout, and CFD?
There are some similarities to DCIM tools, and the visuals do look a bit like CFD, though we are neither. The EkkoSense Platform is a real-time view of power, space and cooling (rather than theoretical (CFD). Whilst we can, and often do, integrate with asset management (DCIM) tools, EkkoSense is 100% focused on the M&E side of the data center, using advanced analytics, to deliver the removal of Thermal Risk, freeing up of M&E capacity, average savings of 30% cooling energy, early warnings for failures, easy reporting (SLA’s etc.), and a simple, intuitive, management tool, for an entire estate.
Is there a limit of stored data, and how is it stored?
No, there is no limit to data stored. We take security extremely seriously and we have installations at some of the largest financial institutions in the world. All data transfers are encrypted and all wireless data is transferred using 128-bit AES encryption. EkkoSense will also shortly have formal ISO 27001 and Cyber Essentials certification.
How many sensors per rack are you using?
Whilst 80% rack coverage (at 1 sensor per rack) is enough for the advanced analytics to work, we recommend 1 sensor per rack and 100% coverage. If you already have existing sensors, we can use those, if you require sensors, then our own EkkoSensors are low cost, and easy to deploy.
Walk-through the EkkoSense platform and learn how it can be implemented as a part of your data center energy efficiency strategy by watching the webinar.