Flexible, Sustainable Data Centers
Alan Abramson, PE; Oliver Varban, PENG; and Wayne Howell, PE
Sustainable data center design can be a tricky business because of the inherent need for an organized yet flexible approach. However, in recent years, design firms have come up with solutions that provide sustainable and highly flexible designs while also steering clear of confusion related to rapid changes in computer hardware technologies.
The key to ensuring flexible yet organized electrical and mechanical systems for sustainable design centers involves preparation, becoming informed, and planning ahead. In order to be fully prepared and achieve maximum success, design firms must balance all of the business needs involved in the project and then develop and implement a sound basis of design (BOD) and project plan.
Balancing business needs
To meet the requirements for a sustainable and flexible data center design, the engineer first must balance all of the project’s business needs, including the technical requirements, cost considerations, and possibilities for growth.
Technical and operational requirements
The first business need the engineer should fully understand is the project’s technical and operational requirements. For a project as complex as a new data center building, it is important to clearly define these requirements to ensure that systems are designed to meet reliability needs and that adequate space is allocated for future growth. Knowing the technical requirements allows engineers to design electrical and mechanical systems that best meet performance and reliability criteria. Meeting with the owner’s IT staff at the start of the project to determine the IT operational requirements and project specific technical requirements also allows the engineer to fully unlock the potential efficiency savings and begin developing a successful BOD and project plan.
Another key business need that should be addressed early on is cost. When the owner identifies the project budget, it allows the engineer to select the appropriate equipment and systems to stay within this budget allowance. Making these cost-appropriate selections minimizes design changes and allows the project to stay on schedule.
Possibilities for growth
Engineers should also consider the client’s growth needs when planning their designs. For a new data center building with IT requirements that have a scheduled, phased growth plan, there must be a plan for meeting those needs without disrupting the business operation and incurring unnecessary additional costs.
Deciding where to locate the necessary power and cooling equipment is another consideration. Engineers need to explore alternatives for locating the power distribution units (PDUs) outside the data center itself. For example, computer room air conditioning (CrAC) units could be located in a space adjacent to the data center, such as a service corridor or a lockable mechanical room. locating the powering and cooling equipment in the data center can result in a smaller data center footprint with less space for IT equipment.
However, locating equipment outside the data center removes its associated heat loads from the data center itself. It also maintains a more secure data center, as service personnel need not enter the data center to work on this equipment. This can also keep the data center more dust-free, as air filters are not changed out in that clean space.
Building in additional space and provisions for this growth affords the client an easier means to expand. Then in the future, when electrical and mechanical equipment needs to be added to the building, the engineer has another opportunity to work with the owner and IT staff to specify this new equipment. Providing reliable, scalable systems that meet the client’s needs for the initial phase of the project creates a satisfied client who will look for the engineer to provide additional expertise for future expansion phases.
The BOD and project plan
After the business needs of the client have been fully researched and considered, the engineering team can then work to create a successful BOD and project plan. The BOD details the design aspects of the project, and the project plan describes how the engineering team will incorporate the BOD into their design drawings. The engineering team should use the following important design tenets to satisfy the BOD and ensure sustainable data centers that are both organized and flexible:
- Simplicity: Design the facility to minimize risk of failure from overly complicated designs.
- Scalability: The data center must support growth with minimal disruptions.
- Flexibility and modularity: The data center must support new services without a major overhaul of its infrastructure. Modular design allows creation of larger systems from smaller, more manageable building blocks.
- Reliability: The data center design must minimize or eliminate single points of failure and offer predictable uptime.
- Monitoring and control: The ability to monitor and control every piece of infrastructure both on-site and remotely is an extremely important feature for reliable facility operation.
- Technology: Design should incorporate and support the latest technologies.
Data centers that integrate designs based on simplicity are easier to understand and more intuitive to manage. By keeping the design as simple as possible, engineers also lessen the chance for errors, which can affect uptime, performance quality, and operational efficiency.
A highly operational data center must be able to support fast and seamless growth without major disruptions. Scalability enables owners to sustain rapid performance growth while delivering superior results. When scalability is successfully incorporated into the design, as performance increases, the data center’s design can tolerate very short-term changes and adapt to longer term planned growth trends that would be considered chaotic in other facilities.
The engineering team can build significant scalability into many different aspects of a data center’s design. For example, data centers can be created with the ability to increase square footage, expand the emergency power system as well as power distribution to critical loads, and add power panels as needed for additional connected loads. Expansion module sizes should be coordinated to ensure that the new load needs are met with no operational disruption. Finally, a load monitoring system can be implemented to provide information used to predict the need for additional equipment, which facilitates easier growth.
The engineer should be aware that each rise in power demand requires support from the cooling and plumbing systems. The magnitude of the change dictates how many system components need to be upgraded. Scalability and simplicity in design allow easier planning for future growth while maintaining high-performance delivery, ultimately addressing the client’s business needs.
Flexibility and modularity
A flexible data center design accommodates new service offerings without requiring a complete architectural redesign or additional drastic changes. It is easily upgradable and adapts to changing business conditions. At times, these conditions may include making the design cost-effective, and designing a cost-effective data center is dependent on the mission of the center. Since every design decision has an impact on the budget, the data center’s mission, service life, and long-term plans must be clearly understood during the design phase.
Flexible designs and modular designs are closely related because a modular design actually enables data center flexibility. Fundamental design modules are small, often simple units intended to meet a unique purpose or perform a specific function. They are easily defined, and can be easily replicated to meet the necessary scale. Data centers are highly complex facilities by nature, so care should be taken to minimize variations in the designs of specific system modules. Failure to do so adds complexity, and may quickly make the data center unmanageable.
Each major equipment group of the electrical distribution system in data centers can allow for modular expansion. This includes the uninterruptable power supply (UPS) system, emergency power system, and power distribution system.
For example, the power distribution system in the Emerson global Data Center is composed of two electrical rooms, main switchboards, UPS systems, PDUs, foundation distribution systems, ATSs, and power strips in a redundant configuration. By limiting the power distribution to only six primary components, the design team minimized complexity while preserving functionality. The design’s simplicity and modularity provided needed flexibility for data center expansion, allowing the data center to expand its power and cooling up to three times of the original capacity.
High reliability requires a redundant architecture in which failures are predictable and measurable. various standards exist to qualify and quantify reliability, one of which is the TIA-942 Telecommunications Infrastructure Standard for Data Centers. With multiple performance tiers that specify various levels of reliability, Tier III of this standard anticipates 99.982% uptime. This corresponds to an average of only 1.6 hr/year of unplanned downtime. Although this is an aggressive target, the electrical power distribution system in sustainable data centers can be designed to meet this requirement, minimizing downtime and increasing environmental responsibility.
To minimize the number of single points of failure in the electrical power distribution, critical systems such as the electrical service, emergency power generation, UPS, and power distribution for critical load need to be designed with a certain level of redundancy that has been dictated by the owner’s IT staff. Such modular and redundant design allows moving critical loads from one power source to the other without power interruption.
Because rack loads can be quite high, engineers can also consider high-density cooling units mounted directly on the racks for additional cooling and reliability.
Refrigerant is pumped from chillers located in the service corridors to these rack-based cooling modules. These modules draw hot air from the hot aisle, cool it, and then discharge it in the cold aisle.
In order to ensure the rack cooling loads are met, computational fluid dynamics (CFD) calculations should be run. These calculations consider the space architecture, floor to ceiling height, structural members, air flow, rack layout, and locations of high-density loads to determine airflow and space temperature profiles. This enables the HvAC and IT engineers to view how effective the layout is and make any adjustments necessary to achieve optimal airflow for rack cooling.
Monitoring and control
An energy management and monitoring system is critical for added reliability. It provides a way to monitor and control all the pieces of critical equipment, and includes real-time monitoring and control, data analysis and trend reporting, and event management. These functions can be carefully coordinated with and tailored to IT personnel requirements so that they provide the information that is most beneficial to IT. An energy management and monitoring system ultimately gives managers the ability to view information at an instant in the event of a problem. It also gives managers historical information on reliability, behavior, and trending at their fingertips, helping them to be proactive and make the right decisions to avoid future problems, improve efficiency, and fully leverage their investment. This type of monitoring also is useful for energy management, allowing the measurement and verification necessary to meet lEED requirements.
The data center should be designed to allow for easy incorporation and support of the latest technologies and products. Many new products improve operation of the data center with better communication capabilities and data as well as energy usage monitoring functions that provide valuable information to the IT and building engineers. Having a flexible, scalable design plan allows the data center to easily incorporate the continually evolving technology that is a hallmark of IT systems.
Today’s data centers are critical multimillion-dollar facilities. Without them, business as we know it could not exist. However, effectively designing these centers to be sustainable, scalable, and deliver outstanding performance—while avoiding unnecessary complexity in the design—is a challenge many IT departments face. By planning ahead and fully addressing the business needs of the project and then developing and executing a strong BOD and project plan, engineers can successfully design sustainable data centers that are flexible, highly energy efficient, and organized.
About the Authors
Abramson has more than 35 years of experience in the planning, engineering, construction management, operations, and maintenance of buildings. He currently serves as president of Clive Samuels & Assocs. Inc., an Emerson Climate Technologies company. He has designed and managed mission critical facilities for major corporations and financial services companies. He also has served on the adjunct faculty of the New York University Real Estate Institute, the new york Institute of Technology, ASHrAE, and the APPA Institute for Colleges and Universities. Varban has more than 30 years of experience in the design and construction of electrical systems for mission critical facilities, and commercial and industrial buildings. He is a lead engineer at Clive Samuels & Assocs. Inc., a licensed professional engineer in Ontario, and a member of the IEEE. Howell has more than 18 years of systems design experience for commercial buildings involving supermarkets, restaurants, retail, mixed-used, institutional, and commercial buildings. He currently serves as principal and chief engineer of Clive Samuels & Assocs. Inc.
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.