The Risks of Lack of Monitoring for African Telco Shelters

There is little to no monitoring in one of our African client’s Telco shelters!  What are the risks and is there potential for cost reductions?

The client in question owns hundreds of these tower shelters that are housing their cell tower and IT equipment.  These are dispersed in various locations throughout the country.  There is currently no operational monitoring at those sites.  There is, therefore, no way to know how well these sites are performing.  Their energy consumption, the status of their backup generator, fuel levels, and potential intrusions are currently unknown.

Continuous monitoring can answer all of these questions in real-time.  With three months of limited data, it was established that site #60 has a fairly good PUE of 1.53.  But the site’s energy usage is greatly dependent on outside temperature due to its low insulation.  Savings of $600 per year could be achieved by adding 6” of insulation to the exterior of the building.  Also, the outside air temperature is cooler than the site’s set point approximately 50% of the time.  This advantage could be harnessed better with “Free Air Cooling” technology.  It was also observed that on two consecutive days, March 28 and 29, between 6 am and 7 am, the cooling unit’s exhaust temperature went from its usual 20C to beyond 26C.  This could be due to something mundane such as maintenance on the units.  But it is an example of the benefits of monitoring.

Continuous monitoring paired with a powerful support system opens doors to many other Operational Expenditure savings.  Maintenance needs can be measured and predicted thus eliminating the need for routine checks.  The need for the capacity increase can be predicted ahead of time and budgeted accordingly.  Energy usage can be measured in great detail and used to accurately calculate Return on Investment for various cost reduction incentives.  All of this makes the overall operation much more cost-effective while greatly reducing risks of downtime.

Cooling Dynamics

IT Loads Vs Utility Load Vs External Temperatures

• SetPoint: 24 deg C

• Ultimate Cooling Capacity: 7 kW

• Available Cooling Capacity: 3.5 kW

• Cooling usage with redundancy: 77%

Observations

• Cooling system capacity is properly matched with n+1 redundancy

• The cooling system is behaving normally with a set point at approximately 24 deg C

• IT Load and outside temperature peak essentially at the same time toward midday.

• On March 28 and 29 from approximately 6 am to 7 am, the exhaust air of the cooling unit went from a typical 20 degrees Celcius to 25 degrees Celcius. These are the only two instances in the 3 months of records. May be worth investigating if maintenance was being done or if doors were intentionally left open for equipment delivery.

Hourly PUE Dynamics

PUE vs Outside Temperature vs Outside Humidity

• Hourly Average Max PUE: 1.56

• Hourly Average Min PUE: 1.22

Observations:

• There is an unusually great variance in the PUE between day and night.

• Since ITLoad and Outside temperature peak approximately at the same time, additional analytics need to be applied in order to determine the causes of this variance.

PUE Influences

Influence of ITLoad on PUE

Influence of Outside Temperature on PUE

Observations:

• By isolating points where the IT Load is fixed, and then applying the same approach to the temperature, the above plots were created to show how these two factors influence the PUE.

• Based on these two graphs, the PUE curve follows the outside temperature curve very closely. The Energy Usage Effectiveness variations are almost exclusively related to the outside temperature.

• We can deduct from this that there is very little insulation in the shelter’s structure.

Calculated Yearly Cost of Cooling

Note: All points taken at fixed ITLoad of 2.6kW

Table 1 – Cooling Costs

Observations:

• The PUE seems to vary greatly with outside temperature (lack of insulation?)

• The average yearly site PUE is 1.53.

Recommendations

Implementation of a monitoring system to greatly de-risk the site while cutting down overall operational costs.

Adding Insulation :

• Based on the observations and with the following assumptions:

• Existing insulation RSI value of 2.6 (3” foam between aluminum sheets)

• average temperature from June to November = 27.3 C (Outside air is warmer than setpoint)

• Addition of 6” of Spray Foam insulation on exterior surfaces, raising RSI value to 8.96.

• The heat transfer from outside will be reduced by 505W

• 505W x 24h x 183 days = 2218kWh x $0.27 = $600 /Year.

Free Air Cooling:

Given that the average outside temperature is cooler than outside temperature about half of the time. It would be worth looking into “Free Air Cooling” units as replacements once the existing cooling units have reached the end of life.