Data center owners want their new facilities to be completed on time, on budget, and operating as intended. This doesn’t seem like a particularly difficult task – except it is. Ask most people involved in data center construction, and they can give you tales of woe about projects gone wrong.
All construction follows a similar path to completion: the owner identifies the need and budgets for a new facility, hires a design engineer, and awards the build to a general contractor and trades who order and install materials and equipment. In the data center world, this construction process is complicated by compressed timelines, complex electrical and mechanical equipment and interfaces, the need for highly skilled trades, and, oh, by the way, a tight budget. The adage says, “You can have it good, fast, or cheap. Pick two.” This statement is based on the project iron triangle of time, cost, and quality; try to reduce any of these factors, and the other two are negatively impacted.
When building multiple data centers, owners may believe that having a standard design refers to the layout and electrical one-line, but it doesn’t. An owner’s “standard” design is typically put out to bid to get the best price. Each project can include a different architect, design engineer, general contractor and trades contractors, and various mechanical and electrical equipment vendors. Each player interprets the owner’s standard design to the drawings, specifications, equipment, and installation. The result is a group of data centers that are not standardized upon completion. When these changes occur from site to site, one building is not exactly the same as another at the most critical level, the operational level.
Have Your Cake and Eat It, Too
But what if you can have all three? HP&D has spent 25 years helping our data center clients negotiate this triangle and traditional construction process, and we’ve realized there is a better way to manage highly successful repeat data center builds. The key is two separate but related and often missing components: standardization and integration.
Key Ingredient – Standardization
Standardization starts with high-level owner buy-in and commitment to consistent thought processes, applications, and solutions. The next step is assembling a team that includes construction, operations, and a standardization/integration specialist to identify critical components and processes while monitoring implementation and enforcement. While there is an up-front cost to develop this level of standardization, the rewards will pay off ten-fold in decreased construction issues and delays, a more efficient, cost-effective commissioning process, and a fully operational data center on day one. If owners skip this process, each project essentially starts from scratch with a combination of new construction team members and equipment vendors, leading to delays, headaches, and costs. With standardization, your commissioning budget is spent in the field, not tinkering and trying to solve problems on the fly. Standardization can also reduce costs by minimizing spare equipment needed across a campus. In the case of a hyperscale site, these savings alone can be significant.
Secret Sauce – Integration
The final and most important step is establishing a common platform (integration) for your communications, including controls, relaying, alarming, and monitoring protocols. Spend time on the front end figuring out how you want your system to look and feel for the operator and find the most points that you can standardize throughout your system. HP&D helps our clients with integration by developing an owner-specific drawing for the switchgear vendors that details how to wire their equipment. We write programs to interface with the controls defining the inputs and outputs for the protection relays and other communications. In this case, you’re identifying, controlling, and standardizing the “brains” of the switchgear, so regardless of the equipment manufacturer, you have a standard design from an operations standpoint. When it comes time to test and commission, you have a known response expectation speeding up and smoothing out the process. Furthermore, from an operations perspective, the outside facing components (labeling, interfaces, alarming, etc.) are identical, creating seamless functionality from building to building.
Many owners mistakenly believe their design engineer is providing this level of service. MEP engineers provide generalized operational requirements, drawings, and specifications, but they don’t provide the details and nuances of integrating all the vendors’ equipment. Most engineers don’t have the knowledge (or the budget) to provide this degree of customization, resulting in a scope gap. Scope gap creates significant issues during construction, the most common being the blame game – whose fault is it that equipment doesn’t function as intended or doesn’t communicate with another vendor’s equipment? This finger-pointing leads to delays and additional cost as there’s often no one in a position to drive the resolution.
The Reward
Standardization and integration overcome the typical construction issues by filling in the scope gaps before the project starts allowing you to bring a data center project to market more efficiently, with higher quality, and generally at a lower price point – project iron triangle success! These keys give each project the best chance of coming in on time, on budget, and fully functional. Everybody wins when a project goes well.
System and Integrations Control Manager Robert Baldwin contributed to this article. Robert specializes in control system analysis, modification, and troubleshooting for various UPS, generator, substation, and automated control systems in data centers and industrial environments.
