Differential Current Measurements in Data Center Environments

There is an increasing demand for solutions that measure differential current in server racks.

The availability of a data center greatly depends on the operational safety of the electrical systems within the facility. A keystone in maintaining operational safety can be described as the monitoring of the differential current. For this reason, DIN VDE 0105 (VDE is a German association for Electrical, Electronic & Information Technologies) specifies the need for regular isolation measurements, which can however be compensated for in ongoing operation by means of differential current measurements.

The consequences of a differential current that is too high may lead to personal injuries, unexpected switching off, or anomalies in peripheral equipment due to the electromagnetic fields caused by a residual current (EMC interference).

What Should Be Measured?
Although many providers of power distribution equipment with differential current measurement emphasize the importance of the measurement, two important aspects are often overlooked:

  1. What currents is the device able to measure?
  2. Where exactly should the measurement be made?

Often, the need for a differential current measurement is incorrectly argued for using the idea of required protection of persons (against electrical shock) and is even maintained that a differential current measurement replaces a residual current circuit-breaker (RCCB). However, while an RCCB disconnects a system immediately from the current network upon reaching a specified threshold value, a differential current measurement reports reaching a specified threshold value without disconnecting the system from the input power. This comparison is therefore completely false since, even in the long run, it can be dangerous to not take further technical and organizational measures, such as: potential balancing and protective grounding measures, system access restriction, workflow management, detailed alarm and emergency management.

What Current Can Be Measured?
Principally, there are two different measurement types used in practice:

  • Type A according to IEC 60755: Measurement of alternating and pulsating direct currents
  • Type B according to IEC 60755: Measurement sensitive to universal current (measurement of direct and alternating current – that is, where all current types are measured)

Switching power supplies installed in servers operate according to the principle that they rectify the alternating current and where the direct current is chopped with a very high frequency. With this method, residual current may arise which contain residual current may arise, which contain DC current or AC current with high frequency. A residual current of this kind can only be detected by devices that measure according to type B. Thus, it is important to use devices that measure the differential current with type B measurements – which are sensitive to universal current.

A differential current measurement must be fully planned. It is not enough to install the devices, for example, in the input circuit of a rack PDU. A measurement system implemented in this position only monitors the residual current after the transducer – i.e., only the total residual current of one part of the PDU and the devices connected to it. The PDU input cable and internal wiring in front of the transducer are not monitored and residual currents occurring are not detected. Essentially, this means that a potential hazardous risk is lying undetected.

The correct way to measure residual current
For the reasons outlined above, monitoring should start at the beginning of the power path – for example: after the UPS.

A type B measurement device – which is sensitive to universal current – positioned at the beginning can, for example, continuously monitor a number of racks, including rack PDUs and the consumers connected to it. Monitoring of this kind is necessary where safety aspects are concerned, since all lines and devices after the measurement device (including the rack PDUs and connected devices) are continually monitored.

In the event of an incident: how can you find where the fault lies?
In order to isolate any potential differential current from occurring, at rack level, for example, additional monitoring devices must be installed directly at the point of transfer within the rack PDU (this is inserted after the back-up fuse- not in the PDU). If your intentions aim at isolating the device or server, you must measure the differential current directly at the output of the PDU (i.e., for each output). A type B measurement fixed installation of equipment would be considerably expensive in the data center environment.

Conclusion

  • Availability and safety are two key aspects in data centers. For this reason, monitoring of the differential current is very important.
  • The differential current should always be measured according to type B
  • Monitoring must be consistent throughout the circuit in order to achieve actual implementation of a requested isolation measurement that is in operation. This means that in all cases, measurement must begin at the source of supply. Depending on the precision of the isolation, monitoring may be implemented down to device level.
  • Continuous monitoring for each rack row makes the most sense in practice.
  • To isolate a fault at the rack level, the current must be measured before the rack PDU – i.e., within the sub-distribution system. A cost-effective solution to achieve this is to use a suitable clamp meter for measurement. Cable routing should support the requirement by allowing for easy accessibility.
  • Isolation at the device level is very expensive. The differential current must be measured according to type B at each PDU output – this implies high procurement costs in addition to increased effort in the execution of regular functional tests.

The implementation of differential current metering as describe above will increase the reliability and operational safety of your data center.

Share This Post:

Steven Bartoszewicz