The importance of power reliability
In this roundtable discussion, data center executives discuss the vital importance of power reliability.
By Brian Phelan, ASCO Power Technologies, Florham Park, N.J.
To gauge the role of power reliability in data center design and operation, ASCO Power Technologies sponsored a national roundtable so data center executives could share their insights and concerns on the topic. The news is sobering.
The drive to achieve and maintain ever higher levels of power reliability permeates nearly every facet of data center design, reported the data center experts. The panelists were executive-level representatives of Microsoft, Morgan Stanley, Deutsche Bank, Digital Realty Trust, and others.
Mike Manos, who was general manager of data center services for Microsoft at the time of the roundtable, and is now senior vice president of Digital Realty Trust, began the discussion. “Companies will continue to drive toward heavy reliability and heavy redundancy requirements.”
Greg Sawyer, a project engineer with Burr Computer Environments, said, “We are seeing more redundancy within electrical and mechanical systems. Clients are becoming more educated regarding the desire for more reliability and levels of redundancy.”
Other roundtable panelists agreed. Redundancy and reliability go hand in hand.
For example, Steve Emert, engineering team leader with Rosendin Electric, said, “Today, every business wants to be up and operating online on a continuous basis. It's been a paradigm shift as to how important data centers are to the success of a business.”
Joseph Soroka, president of ZDT (Zero Down Time) Group for Total Site Solutions, said, “The need has always been there, but as computing power grows and the amount of data being processed grows, the need for greater power reliability is also needed.”
Bob Schuerger, principal, reliability analysis corporate lead for Hewlett-Packard, said, “The biggest change in the paradigm of the data center world that we've seen in the last couple of years is a real shift toward paying attention to the economics of efficiency and the cost of redundancy.”
Redundancy, indeed, can escalate costs quickly. Brian Schafer, director of business development at Highland Assocs., cited an example. “In New Jersey, you want to have feeds coming from two separate substations. Ideally, you have electrical services from two separate utility grids, but that rarely happens. These types of costs are key factors when planning and managing these projects. They need greater power reliability for the densities and the wattages per square foot that these big centers are demanding.”
Sudhir Kalra, executive director, global head of engineering and operations management for Morgan Stanley, said, “After the deregulation process in the electrical utility industry a few years ago, at least in the New York area, we've seen the reliability of the electrical grid actually go down a little bit. It just seems that electrical outages, albeit they may be minor interruptions as the utility company calls them, are happening more and more often.”
Command and control strategies
ASCO asked what's required versus what's desired when implementing command and control strategies for power systems.
Emert started the discussion, and said, “There are two scenarios we see. One is where the owner has developed a fairly knowledgeable and experienced staff in-house and wants those folks to take responsibility for the operation of the systems.
“The other scenario is where the in-house staff does not have the experience and owners are asking for higher levels of automation within the distribution system itself.
“Nobody I know wants to go remote. There are a number of issues involved.”
Ted Martin, vice president of operations for Digital Realty Trust, agreed. “I've never been comfortable with remotely controlling any critical piece of equipment,” he said. “I like employees going to a site first to ensure that the alarm that you did receive is an actual alarm and not a false alarm.”
Kalra said, “From a command and control standpoint, first and foremost you need to get full status information, whether it be the status of the electrical distribution system, power system, generators, UPS, whatever else. Typically, every building, at least in the New York area, has a building or engineering control center, which is manned 24/7.”
Shafer emphasized the importance of having personnel on-site. “People like to physically see and touch the equipment, to be there,” he said. “Ideally, you still want to have a body in control of actually witnessing or seeing anything. If you have to start the generator, somebody needs to be there to make sure that it gets done. That procedures and sequence of operations are followed. That there are checks and balances. But, can you do it remotely? Of course.”
Sawyer looked at control strategies from two perspectives. “There are really two schools of thought with desired control strategies,” he said. “One is to keep it simple, without compromising desired tier redundancy, so end users can easily operate and maintain the system.
“The other is to provide the highest level of redundancy and manual operation possible to reduce the risk of component failure.”
Schuerger said, “That one's really customer-driven. Occasionally, a customer wants to look in remotely and control it. There are people who do it all the time. The utility world in particular has been doing remote control and remote operations for a long time—25 years.”
IT and facilities responsibilities
ASCO then inquired about IT and facilities decision makers. Who's responsible for what vis-à-vis centralized backup power versus distributed power?
Soroka started the discussion. “Our company tag line is 'Bridging the gap between IT and facilities,' which actually refers to the gap between the two departments—IT and the facilities groups,” he said. “Even today after talking about the importance of eliminating this gap of information between departments, we still see it.
“However, there is a noticeable improvement or elimination of this gap in some companies.”
Evangelos Stoyas, a consultant with Power Systems Consulting, said, “IT now pays more attention to facilities personnel as data centers become larger. Who's responsible? It's a joint responsibility.
“IT says, 'This is how much downtime I can accept.' Facilities personnel get the engineering done to achieve that, be it changing raised floors, PCs, decreasing harmonics, or some other aspect.”
Emert said, “This is the biggest war in the business today. This is where the rubber hits the road as to who is responsible for what. If the IT and facility managers are in conflict, then you've got problems. The only way they're going to be successful is to work together on these issues and understand who's doing what and where the demarcations are in the center.”
Kalra went into even more detail. “IT people always feel facilities personnel take too long to get things done and their work costs too much money,” he said. “Facilities staff always feel IT doesn't have a good handle on things. They don't know what the projections are. They don't know where they're going to be in terms of power and cooling requirements, two months or even two weeks down the line. That is a problem.”
Sawyer added another point of view. “IT personnel's biggest responsibility is to ensure their data center doesn't go down,” he said. “In some companies, facilities personnel don't totally comprehend what the impact is on a company if a data center goes down. They may treat the data center as just another portion of the facility.”
Martin took it a step further. “People feel possessive of their turf and don't like other people scratching around in their litter box,” he said. “I understand the concerns that drive that mindset, but I strongly believe that everybody needs to drop the blinders and look at the corporate good, instead of individual group requirements or needs. Sometimes the hardest thing is just getting people to work together.
“Oftentimes, it's just a matter of having the director of real estate and the director of IT sit down and map out a collaborative strategy and say, 'Here's our common strategy to get it going forward. This is how we're going to do it and we're going to make sure our teams do it.' Staying committed to that plan and leading by example can solve all of that.”
Manos summed up the discussion on IT versus facilities responsibilities. Speaking of his experience while at Microsoft, he said, “I think it almost breaks down into an organizational structure challenge. In our organization, I have all the facilities and IT people reporting to me. That has allowed me to build a lot of the synergies that we just talked about. I think you see there are very specific camps. There are the facilities camps and the IT camps, and broadly across the industry you find that that is exactly the case.”
Reconciling green differences
ASCO segued into another hot topic among data center managers and asked the panelists: How is the industry reconciling increasing power densities and the pressure to reduce a data center's carbon footprint?
Schuerger led the discussion. “More and more of it is being mandated by legislation,” he said. “It's no longer a choice whether the customer wants to buy the more expensive one up front or not. He doesn't have a choice.”
Manos offered more detail. “Driving efficiency around power is key. When you think about the upfront costs, it's about 11% of the total cost of the life of the building, depending if you go 16 years, or 20 years, or however long you state the life of the building.”
About his experience at Microsoft, Manos said, “Power costs, especially in the larger facilities that we're building, multi-megawatt facilities, are pretty significant. It's been proven that the total cost of ownership of one new one-rack unit server is now down to a point where the lifetime power costs far exceed the cost of the one new server. So, regardless of if you're depreciating on a three-, four-, or five-year cycle, your power costs exceed your initial capital costs of even the server equipment.”
Schafer said, “Reducing the carbon footprint to satisfy the green environment these days, that's a mission that is really challenging us in the data center world.”
Glen Neville, director, engineering for Deutsche Bank, said, “When you're designing your data center environment, you obviously want to be as energy-conscious as possible and try to reduce energy consumption to the extent possible and look at innovative technologies during the construction.
“But after that, we're looking at various ways to reduce our carbon footprint, from installing solar in certain locations where it's cost beneficial, to reviewing the options for purchasing renewable power from some of the vendors in the marketplace to try to reduce our carbon footprint.”
While the executives' wide-ranging discussion addressed several aspects of increased power reliability for data centers, the common theme throughout was that business-critical continuity for their operations remains a “must-have” item.
Phelan is director, marketing services, with ASCO Power Technologies.
The panel of experts
ASCO Power Technologies recently sponsored a national roundtable to discuss power reliability issues facing data center managers today. Ten executives from Fortune 500 companies and industry leading design/build firms shared their expertise and experience on issues including the most recent data center trends, command and control strategies, and reconciling power density versus reducing carbon footprint, to name a few.
Steve Emert %%MDASSML%% Engineering Team Leader, Rosendin Electric
Sudhir Kalra %%MDASSML%% Executive Director, Global Head of Engineering and Operations Management, Morgan Stanley
Michael Manos %%MDASSML%% Senior Vice President, Digital Realty Trust (At the time of the roundtable, Mr. Manos was with Microsoft Corporation as general manager, data center services. It is his experience while at Microsoft that he shared during the discussion.)
Ted Martin Vice President of Operations Digital Realty Trust
Glen Neville Director, Engineering Deutsche Bank
Evangelos Stoyas , PE Power Systems Consulting Engineer
Greg Sawyer , PE Project Engineer Burr Computer Environments
Brian Schafer Director of Business Development Highland Associates
Bob Schuerger , PE Principal, Reliability Analysis Corporate Lead Hewlett-Packard
Joseph Soroka President of ZDT (Zero Down Time) Group Total Site Solutions
Case Study Database
Get more exposure for your case study by uploading it to the Consulting-Specifying Engineer case study database, where end-users can identify relevant solutions and explore what the experts are doing to effectively implement a variety of technology and productivity related projects.
These case studies provide examples of how knowledgeable solution providers have used technology, processes and people to create effective and successful implementations in real-world situations. Case studies can be completed by filling out a simple online form where you can outline the project title, abstract, and full story in 1500 words or less; upload photos, videos and a logo.
Click here to visit the Case Study Database and upload your case study.