Cogent Communications Miami Data Center

EET validated CryoGenX4 at this location and proved an 18.9% savings.

Testimonial:

“Energy Efficient Technologies deployed CryoGenX4 in five Cogent Data Centers. There was no downtime required in our critical cooling systems, we saw the savings in the electric bills and did not require any additional maintenance afterwards.  I highly recommend CryoGenX4!”

 Rich Rademaker, Director of Data Center Operations, Cogent Communications

On 17 June, Energy Efficient Technologies (EET) and Tri-S
Technologies (Tri-S), with logistical support from Cogent
Communications, began an HVAC performance improvement
demonstration at the Cogent Communications data center location on 1st Street in Miami, FL.

Detailed baseline and post measurements were taken of two computer room air conditioning (CRAC) units at this location. The primary purpose of this exercise was to demonstrate the financial savings, increased efficiency, and carbon footprint reductions achieved by implementing the energy control measure (ECM) CRYOGENX4.

The location for the demonstration was selected by Cogent management. It is a single contiguous room with three ~20-ton CRAC units, all of which were treated with the ECM, providing the cooling needs for the facility. Two small, portable coolers are also being utilized in the hottest part of the facility. They were not treated with the ECM.

Four different testing methodologies were utilized during this pilot to illustrate the improvements brought about by CRYOGENX4. Each is described below and expanded upon in subsequent sections of this report.

1. IPMVP (international performance measurement and verification protocols) measurements of
CRAC’s 1, 2, and 3 in the baseline and post periods.
2. Continuous power meter monitoring of CRAC 2 before and after the introduction of ECM.
3. Continuous supply & return air monitoring of CRAC 2 before and after ECM introduction.
4. Continuous monitoring of the Data Center temperature before and after introduction of ECM.

Results:

Power Logging Testing Methodology

As a secondary collection methodology, the Tri-S team logged power usage over a 17-day period. With these 7,680 comparison data points taken at 1-minute intervals, the teams then compared baseline and post power data during similar daily peak load times of 8AM – Midnight for 8 consecutive days in both the baseline and post periods. Using the data in this manner eliminated the overnight dips in temperatures into the low 70’s, mainly in the baseline period. The humidity levels were nearly identical over these periods, while the post temperatures averaged 1.5° warmer. No adjustments were made to the data for weather. The analysis of CRAC 2 power data yielded the results shown in the detailed charts below.

It is important to note in the data that the power usage of CRAC 2 increased in the post period shortly after the ECM was introduced into the system. This is very common as it can take several days of runtime for the CRYOGENX4 to work (I.e., seating period), especially in older systems like this one. Given the ECM was installed at midday on June 27, if you look just at the last 5 days of data (i.e., 1-5 July), the 16.47% reduction in kW usage during that time nearly matches the 17.3% results shown in the IPMVP data. This serves to corroborate both data sets.

Also, in the post power usage chart above, you can see that the kW savings are manifesting themselves from the increase in capacity. CRAC 2 is clearly producing more capacity which is enabling the system to shut-off 9-times in the post period versus only 3-times in the baseline. The durations of the downtimes are also longer in the post. The other big kW savings are coming from the second compressor being able to stay off for longer periods in the post. You can see this clearly with the white lines appearing in the thousands of post data points while there are none visible in the same baseline period.

Data Center Temperature Logging Methodology

The fourth and final method used to show improvements generated by the CRYOGENX4 ECM was to collect the temperature inside the data center over the duration of both testing periods. A recently calibrated temperature logger was placed in the approximate center of the large contiguous room approximately 7’ above the floor with unobstructed airflows in all directions.

As the data clearly illustrates, despite the slightly warmer weather outside in the post period, the average temperature inside the data center dropped nearly a full degree over the testing period.

Other Data Center Results: