Businesses are increasing the amount of due diligence completed before greenlighting the acquisition of new data center space. They are wise to do so. The options available to run business applications and maintain continuity have increased in recent years. There is a dizzying array of cloud and colocation options available. The days of simply increasing floor space at an existing enterprise data center are winding down, too. Earlier generation data centers are either fully “built out,” have become functionally obsolete or have been expanded to the extent possible. When evaluating expansion options or new construction, customers are finding that many existing data center candidate facilities and sites have either leased-up or been acquired. Surrounding properties are not as easy to acquire as they were during the great recession. And the low cost of capital also has many customers evaluating build-to-suit options in concert with tech refreshes.

The following offers the reader an approach to making sense of their options.

SEGMENTING OF BUSINESS OPERATIONS

Many firms are questioning whether a new bricks and mortar data center solution is necessary. Businesses should assess and segment each of their business applications. The majority of those that are undertaking these assessments are developing strategies that leverage a mix of solutions. Few are placing all applications in their enterprise data center or all in the cloud. They are arriving at a decision to engage each of these offerings as it best supports their capital availability and the need for flexibility while aligning with their business needs.

• Cloud. Email, human resources, back office tasks, project related data, core business, accounting and finance, trading, credit card processing, training, etc., each of these business segments lends itself to evaluation of criticality, security, redundancy, and disaster recovery. Through this process a number of companies are finding that cloud solutions, such as Microsoft 365 or Google Enterprise Applications are suitable for desktop applications and other cloud solutions, like Box or Dropbox are suitable for data storage. Similarly, Amazon’s Web Services provides a broad variety of variable, on-demand IT services. In many instances, businesses are choosing to place certain applications in the cloud, instead of expanding their current facilities.

• Colocation. The definition of colocation has expanded over recent years. Whereas providers initially owned the power and cooling infrastructure and leased space for equipment inside data centers by the square foot, those providers have expanded their offerings. Today these providers segment their offerings between retail and wholesale and offer space to clients seeking a room, pod, cage, cabinet, rack, managed services, or similar configuration with varying levels of services through a variety of buildings representing different levels of redundancy.

• Standalone data center. Whether renovating an existing building or developing a new greenfield site, the standalone data center appears to be the solution businesses choose to host their most critical applications. These highly redundant buildings represent a significant investment and the realization of careful planning, whether a stand-alone enterprise data center for a single user or colocation facility to host others, these highly redundant buildings represent a significant investment and the realization of careful planning.

When the decision is made to develop a new data center, the business will need to evaluate the physical space and power needed to meet the business application requirements for near, mid-, and long-term horizons and then what level of redundancy is needed to support it. Once those requirements are understood, a business will undertake a thoughtful site selection approach.

SITE SELECTION APPROACH

Now that the business has opted to develop a new data center, it needs to engage a detailed process to identify suitable locations, and research and evaluate the suitability of each location. This is also an opportunity to identify internal and external resources needed to complete the process. The business may also choose to engage a variety of outside specialists or trusted advisors, some with local or specialized knowledge, to complete these tasks.

• Scope of physical requirements. Based on business application requirements, the business can estimate the amount of floor space and power needed to accommodate the equipment (racks and servers) to support the strategy. This is best achieved through programming meetings with the various IT and infrastructure stakeholders and your trusted data center advisor. Given the raised floor space requirements and desired level of redundancy, engineers can then estimate the mechanical and electrical support infrastructure required. Similarly, a data center architect can estimate the space needed to accommodate this infrastructure and compliment any administrative needs the building may serve based on the developed building program. The result is an estimated floor plate for the new building, including outside space mechanical and electrical support infrastructure, fuel and water storage, delivery space, and parking.

• Initial assessment criteria. Once the scope of the project is set, now it’s time to detail the criteria by which each potential site will be evaluated. That criteria should include not only the availability of specific features, but also the quality of each. The assessment criteria should also rank the necessity of each of those features based on importance to the business. For example, what is the availability of power and fiber, zoning/permitting, construction costs or incentives? For some businesses, a specific location may be the top priority. Another business may rank cost of power as a top priority. Yet another business may rank speed to market as the top priority. Prioritizing each of these criteria would lead the business to a different location. A partial list of assessment criteria appears later in this article.

• Location identification. Few businesses have the flexibility to locate a data center just anywhere. There is usually some overriding conditions that will drive location. Some business applications may need to be accommodated within very specific geographic locations, trading for example. Other businesses may seek to locate their data center in close proximity to wider business operations. Some may seek to locate in a market that leverages special labor forces. Identify and prioritize any geographic opportunities or constraints.

• Desktop research. A remarkable amount of property data and related information is now available and accessible by real estate brokerages and consultancies online. Once preferred markets are identified, gather detailed information on all viable properties prior to conducting site visits.

• Site visits. Despite the availability of online resources, there is no substitute for visiting each target property and gathering information necessary to complete the initial assessment criteria.

• Short list. Consolidate all of the initial assessment criteria and score each of the subject properties. A short list of three to five properties usually provides the owner enough flexibility and leverage to reach a final decision.

• Final assessment criteria. Once the short list is created, develop a more detailed list of criteria that will help your business find the subject property to be meet your business needs.

• Detailed investigation. This is the stage which the business works through the details of realizing a fully functional data center at each subject property. The renovation of an existing building would naturally be thought to yield a quicker speed to market, but upon further investigation the total timeline for development could take longer than a greenfield build. Perhaps there are two properties located in the same community but are zoned differently. The first property is appropriately zoned to allow a permitting process of a few weeks. The other property could require re-zoning, a process that could take months.

Detailed review, recommendation, and vetting. Upon completion of the detailed investigation criteria, consolidate and score each against the pre-determined weightings. These steps will typically lead to the identification of the subject property that best meets the business’s requirements. Even so, it is prudent to vet the findings with the internal and external experts prior to providing instructions to the real estate professionals on the team. This is also the time when the business will develop a more detailed budget and schedule for the project.

SITE SELECTION CRITERIA CONSIDERATIONS

Selection criteria can vary widely, depending on the goals of the project. The following list identifies the major categories that are most frequently considered when evaluating subject properties. Ideally the criteria develop into a matrix. Acceptable parameters are set for each criteria. The criteria are then weighted and subsequently ranked based on priority.

· Physical space. The existing building (renovation) or raw land parcel (greenfield) needs to be sizable to accommodate the proposed design.

· Business climate

• Labor. Availability and quality
• Incentives
  •Tax incentives – sales, property, income, training, employment – job credits
  • Peak shaving power incentives
  • Cogeneration incentives
  • Utility rebates
  • Free air cooling

· Real Estate

• Land
  • Out of 500 year floodplain
  • Topo (minimal balancing)
  • Shape/layout
  • Accessible, etc.
   
• Zoning and permitting
  • Data center friendly zoning
  • Data center friendly build permit review process
  • Expediting
   
• Real estate incentives
  • Land grants
  • Development authority assistance
   
• Acquisition strategy
  • Purchasing entity
  • Type of contract – purchase & sale agreement (PSA), options contract, LOI
  • Own vs. lease (analysis)
• • Control – sale-leaseback
   
• Development model selection
  • Current owner willing to sell?
  • Controlled by a developer wanting to develop?
  • Partner with a developer or self-develop?
  • Process for choosing the developer or general contractor
  • Each model allocates risk and control differently

· Utilities – power. The availability, quality, and cost of power is usually one of the most important priorities in determining suitable locations for a data center.

• Availability of power. Most businesses are seeking redundancy of the critical commercial service to the property. Diverse feeds from separate grids/substations are ideal. Similarly, it is preferred that those diverse feeds enter the property through separate entrances on opposite sides of the building. Increasingly, businesses are also seeking green power options for their facilities. These utilities may derive their electricity from a variety of sources including hydro, wind, or solar sources.
• Reliability of power. The reliability of a commercial power grid is as important as the availability of the system. There are two widely accepted indicators for the reliability of these systems; the system average interruption duration index (SAIDI) and the system average interruption frequency index (SAIFI). High-reliability of the physical structure (buried in concrete encased ductwork) is desirable due to reduced overhead exposure risk.
• Cost of power. The cost of power is often the single biggest expense line item in the long-term management of a data center, as well as TCO (total cost of occupancy) modeling. Businesses typically evaluate a variety of options to weigh power costs that can include inputs from utilities, engineers, and the local environment including; negotiated rates with local utilities, available power rebates or incentives and free air cooling strategies. Power costs affect both the connected load and the back-of-house support infrastructure charges — power utilization efficiency (PUE).

· Utilities – water. Availability of water is a priority, it is receiving increased attention as a result of regional droughts that have taken place over the last decade. The Southeast experienced a drought from 2007-2009 that caused concerns. Some of that concern has now shifted to the West, which is currently experiencing a drought. Most businesses are not seeking the same level of redundancy from their commercial or well water supplies, as they do power. This is largely because short and mid-term water supplies can be easily stored on-site. Nevertheless, availability and vulnerability of the water supply is important and must be evaluated.

• Availability and reliability of water supply. A vulnerability assessment of local and regional factors that could impact availability is recommended. Factors can include natural events like drought, hurricane, or earthquake or adversarial actions like vandalism or terrorism. Businesses are also likely to evaluate the potential of local restrictions for water use. Utility delivery design is evaluated with looped supply both in the street and eventually around the site preferred to increase reliability. Many of these factors can be mitigated through procedural or policy changes to mitigate the risk.
• Cost of water. Businesses typically evaluate utility delivered, surface, and hydraulic options for water supply and the availability of incentives or rebates.

· Connectivity. Equally important as power is the availability and quality of connectivity options —  fiber providers.

• Availability and reliability of connectivity. Most businesses seek redundancy of their fiber connectivity backbone. Diverse feeds from different providers originating from different paths are ideal. Similarly, it is preferred that those diverse feeds enter the property through separate entrances on opposite sides of the building.
• Cost of connectivity/bandwidth. Another significant operational cost is connectivity. Similar to hosting options, there are a variety of connectivity options offered by increasingly sophisticated service providers. Most businesses actively seek to locate their data center in an area where there are multiple service providers, to ensure competition. Many enterprise data centers will bring more than one backhaul provider into the building. Colocation operators will usually bring multiple backhaul providers to the meet-me room(s) in an attempt to provide their customers with the best combination of availability, reliability and cost.

· Physical environment. The chosen location and immediate surroundings represent some of the most significant factors in events that could disrupt operations. The following list identifies the most frequently evaluated factors.

• Natural disaster; seismic, tornado, hurricane, flood, wild fire, tsunami
• Hostile adjacencies; airports, railway lines, nuclear, hazardous/explosive materials
• Adjacency to industrial or residential neighborhoods; power quality issues and noise may impact the surrounding neighbors.
• Highways. Businesses want to locate close enough to transportation options that provide for the convenient flow of materials, equipment and staff, while concurrently avoiding being impacted by a traffic accident or similar event.

· Physical structure. The physical structure of the building containing the data center provides the greatest opportunity to mitigate many of the risks identified in this article. It is also the piece where the business can have the greatest control and influence.

• New or existing. Unless seeking a highly unique data center solution, most businesses compare the renovation of an existing building and the development of a greenfield site. During and shortly after the recession there was a trend towards renovation due to the surplus of suitable buildings. Many of those most suitable buildings have either already been converted or repurposed for other uses since the economy has improved.
• Steel or reinforced concrete. The structure of a data center building is typically steel, reinforced concrete, or similar material.
• Single story, two-story, or multi-story building.A single story building provides the best opportunity to mitigate overhead risk. Two- story buildings allow for direct servicing of a second floor data hall from the first floor infrastructure. The authors acknowledge that a great many data centers are located in multi-story buildings for various reasons.
• Flooring to support load. Data centers require floor loads that exceed those of typical commercial construction. Evaluation of this requirement is especially critical in the renovation of an existing building or in multi-story configurations.
• Overhead structure. In a single story configuration, it is essential the roof is constructed to withstand unexpected weather (external) and support hanging loads (internal). In multi-story configurations the evaluation of the overhead structure will require the cooperation of the tenants above. Don’t forget the hanging load from the ceiling of one story needs to be combined with the floor load of the story above.
• Clear space. The need for raised floors and space for overhead racking necessitates significant overhead height.
• Minimize wall and roof penetrations. Wall and roof penetrations represent risk for water intrusion. A design that minimizes these penetrations is ideal.
• Physical barriers. Many businesses site the building on the property to minimize the risk of physical intrusion to the facility. Other businesses take further measures to install physical barriers to limit unauthorized vehicular and pedestrian access to the facility.
• Generators. The site plan must include considerations of space (near and long-term) for generator and fuel capacity sufficient to drive the building’s infrastructure. Further considerations should include that local noise and environmental restrictions are met.
• Mechanical. The site plan must include considerations of space (near and long-term) for the mechanical system and water storage capacity sufficient to drive the building’s infrastructure.

In conclusion, the process of site selection — done right — involves considerable assessment, planning and due diligence.  The significance of this process has been amplified by the varying and increasing options available to data center occupants. The timeless advice to "measure twice and cut once" certainly applies. Employ the expertise of specialists to strengthen your team; IT consultants, attorneys, Real Estate brokerage advisors, engineers, and contractors.