Sarju Tailor is Senior Product Manager for GE’s Critical Power Business.
As Einstein taught us, measuring anything is relative. In the data center arena, we measure processing speed, computing capacity, data input and output and demands for storage — over months, weeks, days and even seconds. Yet, when we measure the cost of acquiring and operating our data center equipment, we too often select a single point in time to measure cost. We calculate one-time, up-front capital expense (CapEx) and initial start-up operating expenses (OpEx)such as power equipment’s rated energy efficiency.
Calculating total cost of ownership (TCO), however, begins by looking at the long-term view, well before equipment is even selected and ending far past the operational life of the system. Looking at TCO for data center power systems, particularly uninterruptible power supply (UPS) units, offers a vantage on how to evaluate TCO across the full life cycle of data center systems. For example, typical TCO metrics include single point-in-time TCO factors such as UPS power conversion energy efficiency. This is certainly a key OpEx consideration, but when looking at CapEx and OpEx across the full life cycle of a UPS system, a number of other seemingly small but interesting TCO variables emerge.
Defining a Life Cycle
The first core question to ask is, "how long is the life cycle of the system that is being measured?" Is the life cycle based on the practical use-life of the data center system or, more often, just the UPS design life cycle? Defining the practical use limits of a system affects both the CapEx amortization schedule and long-term OpEx and maintenance expenses that are vital for an accurate UPS TCO evaluation.
The TCO of Evaluation and Selection
When beginning the evaluation and request for proposal process for a UPS system, businesses factor in the upfront design and specification time by typically engaging third-party consulting engineering services to handle the design, specification and evaluation of power protection providers. Part of the evaluation process is testing and maintenance costs associated with large or multi-location deployments. These points all factor into how to amortize the cost of the system over the life of the power protection system.
Deployment and Infrastructure
Most data center engineers and managers understand the costs related to deploying new or upgraded power protection systems. Certainly the time and staff costs for installation and initial testing are traditionally part of the overall cost evaluation and TCO metrics. Yet, there are often some hidden costs that also need to be factored into the TCO calculation. During an upgrade or expansion project, do some systems have to be taken off-line, and is the cost of this downtime factored into the TCO? Have new or expanded power distribution infrastructure issues such as expanded or redundant wiring been factored into the overall system cost and TCO?
As data center floor space is a fixed—and valuable—asset, it too becomes a major consideration for calculating overall TCO. Creating a smaller power equipment footprint means more space for the server cabinets which ultimately increases processing capacity and earns revenue.
Maintenance and Support
The costs related to ongoing maintenance, warranty costs, etc., are common metrics for measuring TCO; however, there are some less obvious maintenance-related costs that also need to be rolled into TCO evaluations.
For example, does the UPS topology have sufficient redundancy that allows a single UPS unit to be taken off-line for maintenance or evaluation, or does the entire power plant need to shut down while maintenance or repair is performed? Even scheduled maintenance has an effect on uptime, data and processing transfer time and costs, including labor costs.
Scheduled battery replacement is probably the major OpEx cost of a UPS, representing a significant part of a maintenance budget. If TCO is a critical evaluation factor, then understanding which battery technologies can extend the life cycle of a UPS becomes important. The same is true for remote UPS monitoring systems that improve battery life, maintenance and upgrade strategies.
Lastly, it’s important to understand that maintenance costs escalate over time. Just as an older car costs more to maintain, older UPS units require more maintenance, skewing later OpEx costs and ultimate TCO evaluations.
What are the end-of-service costs in terms of decommissioning data center power infrastructures, and what are the hard costs (e.g., battery disposal) and less-defined carbon-impact costs of disposing or recycling power products? Is there a net-even valuation of batteries, for example, for decommission costs versus recycling value of the core components and materials?
So as we look at important CapEx and OpEx factors to calculate data center power UPS systems TCO, we need to be diligent in understanding and fully measuring these costs at points along the system’s life cycle. From the earliest evaluation steps, across operational and maintenance costs and through end-of-service factors, an accurate TCO analysis demands this insight.
Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library.