Data Center Trends: Driving The Data Center

(Photo: Thinkstock Images)
(Photo: Thinkstock Images)

By Kristy Rainwater
From the March 2013 issue of Today’s Facility Manager

Business owners and directors across numerous industries are beginning to acknowledge that the day-to-day operation of their data facilities is just as critical as the initial engineering, design, and construction of these assets. This growing trend is evident in the design development phases, as owners are taking a greater interest in how their facilities will operate once construction is complete. Human error and the costs associated with resultant downtime, coupled with the rising costs of energy and substantial opportunities for reducing energy consumption, are driving owners to take a closer look at how their data center assets are managed and operated.

The popular saying, “you can’t manage what you don’t measure,” has rarely held more meaning than in the data center environment. It is also essential that facility managers (fms) take a holistic approach to maintenance and performance. As a result of more detailed awareness and control of facility resources, rising energy costs can be mitigated. This is critical, especially when taking into consideration that energy costs can account for approximately 70% of data center operational costs according to a Forrester Research Report (Updated Q3, 2011: “Power and Cooling Heat Up the Data Center”).

Enacting a comprehensive management plan that enables new or legacy data centers to drive effective day-to-day operations, energy efficiency, cost management, and optimized performance calls for several components. Experience shows that among the most critical are:

  • preventive maintenance;
  • building management systems (BMS) integration;
  • operational baselines;
  • retro-commissioning; and
  • local utility engagement.

Chris Koerner, critical facilities manager of T5@LA, which recently completed construction of a 205,000 square foot data center in Los Angeles, CA, would say it starts with knowing the facility’s equipment—intimately at that. According to Koerner, “Developing a comprehensive preventative maintenance program with equipment specific methods of procedure (MOPs) and extensive documentation is critical to the operation of all IT, electrical, and mechanical data center equipment.”

Identifying repair opportunities before equipment fails not only reduces unplanned capital expenditures, but doing so also reduces the potential for disastrous impact events that customers experience as downtime. Keeping detailed records of the operational nuances of systems and equipment allows fms to detect potential issues sooner rather than later when a reading or measurement deviates from the norm. This level of documentation also becomes valuable when justifying operational budgets or evaluating opportunities to reduce expenses without impacting reliability.

Koerner uses the data center’s BMS to capture and cumulate operating data at three minute intervals (including outside air temperature, prevailing wind, external and internal wet bulb and dry bulb conditions, internal static pressure, and air temperature) throughout the day. “Establishing operational baselines, assessing daily operating trends, and tracking Power Usage Effectiveness (PUE) helps identify trends or irregularities in energy usage and operational performance that, when interpreted proficiently, can be used to optimize performance throughout changing operating conditions and potentially avoid costly downtime,” says Koerner.

Through this analysis he has developed an intimate knowledge of how different weather patterns and internal and external conditions can affect the operation of his equipment. In addition, Koerner captures the facility’s PUE at five second intervals, which enables him to identify opportunities for immediate operational or energy efficiency improvements.


The two steps mentioned above can also serve as the foundation for understanding a data center’s energy profile/usage and indicate when a retro-commissioning project would make good sense. As a systematic process for improving an existing facility’s performance by identifying and implementing relatively low-cost operational and maintenance improvements, retro-commissioning helps ensure that facility performance meets an fm’s expectations. The process includes testing and adjusting building systems to meet the original design intent and/or optimize systems to satisfy shifting operational needs. It relies on detailed building and equipment documentation, along with functional testing to optimize performance.

In essence, retro-commissioning is a whole building tune-up. Similar to a vehicle’s 100,000 mile service, the end goal is to achieve maximum operating performance by primarily employing minor changes, and perhaps identifying further capital improvements and/or energy retrofits. As with any system, operational issues develop over time when components degrade as a result of usage, age, deferred maintenance, and modification. In a data center, where rightsizing system operation is a moving target due to continuous changes in power density and heat load, retro-commissioning can have a huge impact.

This 2010 graph shows the cumulative cost to operate a 2,000 square meter data center in Canada. Illustrated is a 10% annual energy increase, along with the observation that 75% of the data center’s operations costs are energy costs. Figures shown here demonstrate that the facility’s operational costs were three to five times its capital cost. (Credit: ©IBM Corporation 2010)
This 2010 graph shows the cumulative cost to operate a 2,000 square meter data center in Canada. Illustrated is a 10% annual energy increase, along with the observation that 75% of the data center’s operations costs are energy costs. Figures shown here demonstrate that the facility’s operational costs were three to five times its capital cost. (Credit: ©IBM Corporation 2010)

In the data center environment, a retro-commissioning process targets:

  • deferred maintenance issues and equipment operating condition;
  • proactive maintenance practices;
  • energy management and sequence of operations; and
  • the relationship between power and cooling capabilities compared to critical load capacity.

This process can reveal problems with very simple or complex solutions. A significant impact can be achieved through minor adjustments or repairs such as:

  • adjusting control sequences that are not optimized or functioning properly;
  • modifying air handler, chiller, and pump staging;
  • repairing or restoring economizer

and variable frequency drive (VFD) functions;

  • repairing or replacing thermostats and/or sensors that are out of calibration;
  • repairing leaking valves; and
  • replacing worn belts and pulleys.

In addition to assessing the operation of existing systems and equipment, further investigation into such processes as air flow management, water and/or air-side economization, and central plant optimization can reveal additional cost saving opportunities with solutions as simple as repositioning computer room air conditioner or air handling units, rearranging raised floor air discharge tiles, adding blanking panels, and increasing temperature set points. Proper retro-commissioning practices can yield energy saving benefits ranging from 12% to 20%, and fms can achieve upwards of that when incorporating energy retrofits such as lighting upgrades and VFDs.

Critical to any retro-commissioning or energy project (and another best practice in general), is establishing a close relationship with local utility representatives. There are local, state, and federal energy reduction programs that offer incentives to organizations for reducing energy consumption. The utility representatives are responsible for managing and promoting these programs that range from direct incentives for equipment upgrades to programs that incentivize data centers for switching to generator power during brownout conditions.

The key to capturing these incentives is to start these conversations early. For most utilities, there is an application processing time that may take anywhere from a few weeks to a few months, and the utility provider may have to conduct benchmarking prior to the start of the project as well, as monitoring and verification follow the implementation of any energy conservation measures.

Aside from the energy savings associated with a comprehensive energy management program, there are a number of other benefits from such plans:

  • increased critical load capacity;
  • reduced operating and capital repair costs;
  • improved operation and maintenance staff productivity, capability and expertise;
  • optimized equipment lifespan;
  • improved computing capacity management—right sized systems;
  • improved asset management/IT deployment; and
  • increased asset value.

According to Gartner Research, data centers consume 30 to 80 times more energy per square foot than traditional office space, and therefore, can tremendously impact profitability. And the changing role of data centers in today’s business environment has prompted C-suite executives to take a closer look at how their IT resources are managed and operated. While energy efficiency is a significant driver, the added benefit of providing users reliability and customers more uptime through operational best practices, is undeniable.


Rainwater is a business development professional for Southland Industries, a national mechanical engineering, construction, and service firm. Specializing in new construction, retrofits, retro-commissioning, energy projects, and facility operations, she is responsible for developing relationships with end users in order to deliver custom design-build-maintain solutions. Rainwater is an active member of AFCOM, Association for Data Center Management Professionals and serves on the Board of Directors for the Southern California Chapter of 7×24 Exchange.