Group 2 medical locations: designing IT systems

Paul Harris, an independent consultant working with the medical locations working group for BS 7671 and author of Electrical Installations in Medical Locations (due to be published by the IET later this year), discusses medical IT system final design.


This article focuses on the specific challenges that may arise when designing medical IT system final circuits for Group 2 medical locations in healthcare environments. It highlights the application of regulations that are located in the general requirements that are important to medical IT circuit design.

A medical location is defined as ‘Group 2’ in BS 7671 and supporting books, such as IET Guidance Note 7, as a:

Medical location where applied parts are intended to be used, and where discontinuity (failure) of the supply can cause danger to life…

For a number of reasons, designers are often faced with the challenge of providing medical IT system final circuits, often referred to as IPS circuits, that are longer than those that are laid out in current standard arrangements provided by manufacturers and associated guidance.

Electrical designers working in the design of healthcare facilities are faced with a number of questions relating to Group 2 medical locations and, in particular, the design of medical IT system final circuits. The questions range from: ‘Can the IT system be run in the same containment system as TN circuits?’ to ‘How far can I run an IT final circuit?’

The latter question is usually based on the fact that there is insufficient space to accommodate the medical IT system cabinet in the necessary area. There is currently a trend amongst design teams to maximise useable floor space within a given footprint. Consequently, in trying to push the boundaries of design, often to the cost of electrical distribution, there is little or no space available for switchgear. This can particularly affect the siting of medical IT System cabinets, which are usually referred to as ’the IPS unit’.

Designers are then challenged as to why this piece of equipment cannot be located in a plantroom on another floor. This often happens, particularly on retrofit projects where the footprint has already been used for many years. In refurbishing the area and accommodating the medical IT cabinet this may mean that a clinical service space needs to be surrendered for use as equipment space, which is obviously not ideal.

Designers should consider the fact that although it is referred to as an ‘IPS unit’, it is in fact a distribution board for specialist circuits with isolation transformer and monitoring systems integral to the unit. Consequently, designers should consider the physical access connectivity between IT system cabinets and the location it serves, resisting where possible the temptation to hide these units in remote plant rooms – i.e. the IT system may be directly above the final circuit it serves but it is a considerable distance for maintenance staff to operate and maintain.

In standard circuit design there are obvious lengths of run limitations, which will determine this value such as voltdrop, disconnection times etc. These values can be corroborated through simple calculation, usually using proprietary software, which usually focuses on maximum Zs values to meet disconnection times. However, as proprietary software does not directly deal with IT systems, an alternative approach needs to be taken.

Looking at the challenges for a designer, they are faced with the fact that a medical IT system is so designed that it does not fail on first fault to earth, so the usual concept of disconnection times for TN systems is not applicable.

Chapter 41 of BS 7671 deals with this by requiring IT systems that do not disconnect the supply on first fault by having warning measures, which is augmented in Section 710 Medical Locations.

Where a supply does not disconnect upon first fault, it effectively means that the system has now temporarily become a TN system. Regulation 411.6.4 requires that an IT system where the midpoint conductor is not distributed (as it is with a medical IT system) is able to meet the following requirements:

      ZsU x Cmin


U is the nominal a.c. rms or d.c. voltage, in volts, between line conductors.
Zs is the impedance in ohms of the fault loop, comprising the line conductor and the protective conductor of the circuit.
Ia is the current in amperes (A) causing operation of the protective device within the time specified in Table 41.1 of Regulation 411.3.2.2, or as appropriate, Regulation 411.3.2.3, for a TN system
Cmin is the minimum voltage factor to take account of voltage variations depending on time and place, changing of transformer taps and other considerations. The value used is currently 0.95.

This has always caused an element of confusion as, technically, with an isolated system there is no Zs. However, in a healthy circuit, considering the following scenario, the term Zs is applicable even if it’s not in the usual context.

We can therefore see that there is an earth fault loop that is comprised of the impedance of:

(a)  the secondary winding (which effectively would be Ze);
(b)  L1 to the point of fault (which would be R1(1) in the diagram);
(c)  L2 to the point of fault (which would be R1(2)in the diagram); and
(d)  the earth path in the diagram, which would be between the points of fault and, in a Group 2 medical location, would be through the protective conductors and the supplementary bonding conductor for that location.

Relating the values back to the standard formula as above, we would find that:

      Zs = Ze +(R1+R2)


Ze     is the impedance of the transformer secondary.

LOR  is the length of run of the cable in metres (m)

R1 = (R1(1) x LOR) +(R1(2) x LOR)Ω

R2= (R2(1) x LOR) +(R2(2) x LOR)Ω

We can use the manufacturer’s data to find Zs, for example:

Z of 10 kVA transformer = 0.158 Ω

Z of 8 kVA transformer = 0.198 Ω

Ia for a 20 A Type B MCB =100 A

Ia for a 20 A Type C MCB =200 A

Consequently, applying the formula in Regulation 411.6.4 for a 20 A Type B device with a 10 kVA transformer gives the following maximum Zs:

     Zs= 230 x 0.95   = 1.0925 Ω
             2 x 100 A

Using the same transformer with a 20 A Type C MCB:

     Zs= 230 x 0.95   = 0.54625 Ω
            2 x 200 A

Using the same transformer with a 20 A Type D MCB:

     Zs= 230 x 0.95   = 0.2731 Ω
             2 x 400 A

Using the 20 A type B results:

      Zs= 1.0925 = Ze + (R1+R2)

Resulting in (R1+R2) = 1.0925-0.158= 0.9345 Ω

If the maximum impedance for any exposed or extraneous conductive part to the equipotential bonding busbar (EBB) is 0.1 Ω (set out by previous guidance) then the R2 maximum value would be 0.2 Ω.

Applying this to the formula:

R1 = 0.9345-0.2 = 0.7345 Ω

As the formula in Regulation 411.6.4 uses the factor 2 x Ia the formula takes into account that there are two circuits involved. The R1 impedance divided by the length of run (LOR) will therefore provide the theoretical maximum cable length (subject to volt drop and energy let-through limitations).

Using the 70 °C copper cable values from IET Guidance Note 1 we can see the following values for  Ohms per metre of cable at a particular size:

2.5 mm = 0.0889 Ω/m, 4 mm=0.0553 Ω/m, 6 mm = 0.037 Ω/m

Consequently, if R1 =using a 4.0 mm2 cable, this would give a maximum theoretical length (subject to capacitive coupling, volt drop and energy let through limitations) of:

Max LOR = 0.7345/0.00553 = 132.8 m

If we use the same equipment but using the 20 A Type C device we find the following:

Using the same transformer with a 20 A Type C MCB:

     Zs= 230 x 0.95   = 0.54625 Ω
            2 x 200 A

Applying the above formula we find that the value for R1 is now 0.34625 Ω.

This equates to a theoretical 62 m maximum length of run using 4.0 mm2 cable.

If a smaller value of transformer is used the corresponding increase in impedance of the transformer will affect the LOR available for the designer to use.

Having obtained a theoretical LOR we still need to satisfy touch voltage requirements of Regulation 710.411.3.2.5, which limits the touch voltage on simultaneously accessible parts to 25 V.

We therefore apply the following formula:

for a.c. systems:       

RA × ld = 25 V


RA is the sum of the resistance in Ω of the earth electrode and protective conductor for the exposed-conductive-parts.
Id is the fault current in A of the first fault of negligible impedance between a line conductor and an exposed-conductive-part. The value of Id takes account of leakage currents and the total earthing impedance of the electrical installation.

Applying this formula we see that with a 20 A Type B device Ia = 100 A.


0.2 x 100 =20 V

A value of 20 V is acceptable.

However, applying this to a 20 A Type C device:

0.2 x 200= 40 V

This is not acceptable, and we therefore need to determine a value to meet the requirement.

RA = 25 V/200 = 0.125 Ω

Likewise for a Type D with Ia being 400 A, the value would be:

RA = 25 V/200 = 0.0625 Ω

As this demonstrates, the standard values indicated in early guidance are provided for pre-determined protective device solutions.

Using the above information, where a design varies from a 20 A Type B protective device, the practice of providing separate protective conductors and isolation washers appears almost counterintuitive in terms of reducing touch voltages (UL).

Designers are to take into account inrush currents etc., which may cause the characteristics of protective devices to be changed from the 20 A Type B devices implied in guidance such as HTM 06-01 A. In those instances account should be taken of the resulting touch voltage UL, which is potentially created by using standard impedance values.

In summary, this article is not advocating that designers should go out and extend all the medical IT system circuits to maximum permissible lengths, as there may be other factors that affect the design and ultimately compliance. Nor is it suggesting that medical IT system cabinets ‘have’ to be located in the medical location, which may compromise clinical functionality.

As with most design scenarios, the design of medical locations is often a balance of a number of factors. What this article intends to demonstrate is that there is no definite ‘one size fits all’ solution. It is the responsibility of the designer to consider all the relevant factors in each individual installation design.