Case study: Retro-commissioning a data center

These examples illustrate how a typical direct M&V process can be conducted in RCx projects.

03/26/2014


Figure 2: Proposed control points to be trended for one of five centrifugal chillers serving the data center.The following case studies illustrate how a typical direct M&V process can be conducted in RCx projects. Also, with advancements in control systems and capabilities of newer BAS both in hardware and software, and more importantly depending on the BAS design and its architecture, the BAS can provide a significant trending capability. In recent years, control designers and control contractors have included some forms of MBCCx in control systems. Taking advantage of the BAS capability for M&V purposes should be mainly part of the design process. Following are some of the measures that can be included in the BAS design process to expand M&V through the building life span.

M&V trending

Figure 3: Proposed control points to be trended to capture actual and setpoint values of temperature and relative humidity on the data center floor.

Under a RCx effort, an M&V plan was developed for a data center designed for a total load of 8 MW. The facility is 120,000 sq ft including data floor, office building, and support areas. The data floors are served by 40 computer room air handler (CRAH) units and designed as hot and cold aisle containment. Figure 2 shows a snapshot of the trend list issued for a single chiller. The trend list includes 18 points for trending. This specific data center has five centrifugal chillers, which results in trending 90 data points for the chillers only.

Figure 3 shows examples of other control points included in the M&V trends for capturing the data floor characteristics such as temperature, humidity, and setpoints. As Figure 3 shows, due to the constraints dictated by the BAS, samples are collected on 5-minute time intervals.

Figure 5: Trending floor-specific lighting electric demand and comparing the results allows detecting abnormalities or higher demand in one space compared with the others. Using this simple comparison could allow BAS or RCx agent to launch further investigation.For the M&V plan to match power usage effectiveness (PUE) Category 3 (as defined by The Green Grid) measurement for a period of four weeks, the total number of data points trended is about 20 million. This example illustrates how the “big data” phenomenon can appear in launching a full-blown M&V during a RCx project.

Based on the field observations, the memory capacity and sometimes network architecture of most older BAS do not allow them to implement a similar type M&V unless one of the following measures is taken:

  1. The measured data points are saved and downloaded in the middle of the trending process to empty the memory space
  2. The number of monitored points in the M&V plan is reduced for the BAS storage to be capable of trending.
  3. The trending interval time is reduced to avoid slowing down the traffic on the BAS network architecture.

Figure 6: This figure shows an isolated benchmarking for a commercial building in San Jose, Calif. This was achieved by using historical utility data analyzing the energy usage on the same weekend day from February through July.Trend analysis

Using the advanced BAS and metering allows engineers to analyze the systems both individually and holistically. However, the challenge is developing an interrelation between factors of each subsystem and the whole system to shape an analytical approach. This form of analysis will allow engineers to step beyond the conventional approach where the total energy consumption of the facility was compared to few variables such as outdoor temperature and humidity or cooling degree days (CDD) and heating degree days (HDD) (Figure 4).

Without knowing the performance metrics of a building, measuring the success of engineered systems or their continuous performance is limited, if not impossible. Some other analytical approaches, such as the one shown in Figure 5, provide exemplary ways for using historical data and analytics to create performance metrics.


Moh Heidari is director of the energy solutions group at AlfaTech Consulting Engineers. He has more than 10 years of experience in energy engineering, Cx, RCx, and energy modeling. He has a master’s degree in mechanical engineering.



No comments