Co-locating QinetiQ’s major EM Test Facilities

 

By Malcolm Rich BEng (Hons) MIEE CEng MAPM,

QinetiQ EM Research and Assessment Services Sales Manager

 

Relocating or building an Electro-Magnetic (EM) test facility is difficult at the best of times, but doing so against a stringent set of requirements covering TEMPEST, Military and Civil EMC would send most people screaming. This was a challenge we took on recently at QinetiQ by preparing for a realignment of our EM facilities and co-locating them on one site.

 

One of QinetiQ’s skills is that it has always been good at doing hard and difficult engineering, part of this has been gained through building our other facilities across the UK. It was also my second time as part of a team building an EM test facility which I consider to be a unique and rewarding experience. So, how did we go about planning such an exercise and what did we learn from our experiences?

 

Our aim for this project was to ensure that QinetiQ could deliver to its customers what they needed at a price they could afford. The following describes the process we went through in developing our new flexible EM facility capable of operating 24/7.

 

The Facilities

QinetiQ’s pedigree can be traced back to many world renowned organisations. QinetiQ was formed from the majority of DERA (Defence Evaluation and Research Agency) which was an amalgamation of such names as RAE, RARDE, RSRE, DGDQA. In the last few years, DERA has been transforming itself into a commercially focused, customer driven organisation which culminated with the birth of QinetiQ and DSTL (see figure 1 below).

 

 

Figure 1: Organisational history

 

At the start of 2002, most of EM test facilities were brought into one business group.  This, coupled with a desire to deliver better value to the customer, lead to a need to rationalise any duplication in the system. The primary EM test facilities brought together were as follows:

 

·   EMC test facility at QinetiQ Farnborough comprising of:

·        EMC test chambers, Mode Stirred chambers, Electromagnetic Pulse Facility (EMP), High Intensity Radiated Field (HIRF) facility

 

·   EMC Test facility at QinetiQ Chertsey comprising of:

·        Large vehicle EMC test chamber

 

·   EM Test facility at QinetiQ Aquila (near Bromley) comprising of:

·        Anechoic EMC test chamber, TEMPEST test facility, Antenna Calibration facility (OATS), Semi-lined chambers

 

·   Marine Approval at QinetiQ Fraser (near Portsmouth) comprising of:

·        Civil EMC test chamber, over water RADAR test range, environmental test facility, OATS.

 

So why move?

The principal reason for moving was the obvious duplication in the support structures and facilities across the company. So, it was decided that we should relocate what we could to one location. The decision as to where to centralise the business was quite easy, completing this task would be more difficult.

 

The site at Aquila is a good example of one in need of rationalisation. When the original buildings were put up at the end of the 2nd World War, Chislehurst was a small village in the countryside surrounded by hop farms and woodland. Today, it is in a highly desirable residential area near to two mainline stations with quick links into London. Therefore, the 20 odd acres of land are extremely valuable and are occupied by a number of businesses which total only 150 people maximum, down from a peak of 1,000 in the early eighties.

 

So, the decision was made to consolidate all of the remaining business groups and relocate it to Farnborough where another EM business group already existed (see figure 2) and there was space available in nearby buildings to put the relocated facilities.

 

 

Figure 2: Part of current Farnborough EM facility

 

What to move?

This was the big question: what to move and how to do it with minimum disruption to customers and work in progress. At the outset it was clear that the Chertsey chamber could not be moved due to its sheer size and the cost this would have incurred. Although parts of Chertsey are now to be redeveloped, we have retained the facility and have just completed the process of having it fully lined.

 

The Fraser facility was relatively easy, we would retain the over water RADAR test range, which we could not perform from Farnborough, and relocate the EMC and Environmental business. We would either use existing facilities or move items easy to transport, such as the climatic test chambers. However, the real engineering feat would be the relocation of the Aquila business.

 

The challenge

TEMPEST and EMC are very similar in that they are both concerned with Electromagnetic emissions from equipment. However, the major difference is that TEMPEST is looking for information within the EMI at lower levels than Military EMC and significantly lower levels than Civil EMC.

 

When the facility was built at Aquila in 1987/88 there was already an existing EMC facility performing military EMC testing up to 18GHz and an old TEMPEST facility of limited quality and capability. The new facility was designed to perform primarily Production Audit Testing on Cryptographic items and would be required to run 24 hours a day and have significantly improved repeatability for conducted and radiated measurements. To do this we had to meet stringent requirements for:

 

·   Radiated noise and accuracy of measurements

 

·   Conducted noise (accurate measurements are a lot easier to achieve for conducted than radiated)

 

This may sound easy, but in reality the most difficult part was the conducted noise through the mains. Reducing this to a level where we could meet the lowest TEMPEST limits required a significant effort on the part of our engineers and those of the contractor at the time, Belling and Lee. It was apparent right from the outset that the current set of power and signal line filters could not supply the correct attenuation or impedance across the frequency range required. This was eventually solved with a bespoke filter design and installation produced by a combined QinetiQ (or DGDQA as it was known then) and the contractor.

 

However, this problem will always occur when setting up a new facility as the noise and impedance of the power supply will cause us problems. To ensure we made it difficult for ourselves, we also wanted the new facilities to be multipurpose to ensure they would be utilised to the maximum. We decide we could not move the TEMPEST screened rooms at Aquila as they were constructed using continuous welded enclosures weighing over 40 tons each, we looked into many novel ways of lifting, cutting, etc, but in the end we knew they would have to be scrapped.

 

However, the military EMC facility and other small test rooms could be moved as they were modular, so all we had to do was build suitable facilities at Farnborough to meet the TEMPEST requirements.

 

Taking shape

Designing the new test chambers to be multipurpose, i.e. not just TEMPEST but covering military and civil EMC, presented us with some very challenging design and installation problems. One particular challenge was the design of the anechoic room to meet the stringent requirements for civil radiated emissions, but also the requirement to go all the way to 40GHz for military EMC testing.

 

By this point we had already appointed a dedicated project manager and a team of technical experts from Farnborough and Aquila to support him. A supplier was selected through competitive tender and work started on the design of this new type of facility and we soon had old problems rearing their heads again. This was compounded by the fact that the TEMPEST limits were classified and special arrangements had to be made so the screened room supplier could have access to them.

 

At this stage our challenges included:

 

·   Resolving the requirements for E and H field attenuation

 

·   Resolving the requirements for power line filters

 

·   Location of the new facilities due to their physical size

 

·   The power supply required

 

·   Access control and security

 

E and H field attenuation

The Aquila TEMPEST facility had a serious problem with H field interference due to a nearby electrified railway line using a third rail for supply, this was resolved by using a particular type and thickness of steel. However, for the new facility we are trying to use COTS (Commercial of the shelf) solutions wherever possible to keep the cost down. To ensure we didn’t boiler plate the design, we performed a series of H field tests to measure the ambient H field noise in the proposed location, this showed that it was not as significant problem as envisaged and would not seriously impact on the design proposed. However, this would only be true as long as the local environment did not change significantly after commissioning of the facility, this could only be controlled through site operating procedures.

 

The E field interference was the other way around, Aquila is relatively quite whereas Farnborough is situated next to an airfield and has approximately 3000 members of staff with all of their associated IT equipment and facilities. However, modern modular enclosures are significantly better than their counterparts from 15 years ago, our only concern was the performance from 18GHz to 40GHz as this is dependent on the quality of doors, vents and fit of panels. This is new territory for the supplier and we are working closely with them to ensure nothing is overlooked, here quality counts for everything.

 

The power supply problem

As anyone who uses a screened room will know, they are usually designed to a maximum power rating based on what will be tested, for our chambers we assume:

 

·   We will be testing power hungry systems

 

·   We will require a variety of power supplies and frequencies, e.g.:

 

·        115/220v/50Hz single phase

 

·        415V/50Hz 3 phase

 

·         DC up to 96V

 

·        Various other power supplies and frequencies including 60Hz and 400Hz.

 

·   The supply has to be relatively clean and free of spikes and surges

 

The first part of the process was to ensure that there was enough available clean and reliable power to supply our new facilities. This was checked by performing power quality measurements over an extended period and by surveying the power supply system to see if our proposed location had spare capacity. Luckily, our new location was right next to a main substation that was lightly loaded, but we still had to contend with the problem of interference and how to reduce it to an acceptable level.

 

One solution we are looking at is to reuse one of the Electrostatically shielded step-down transformers from Aquila, to provide us with our three phases supply for the facilities. Additionally, we are looking to have only the screened rooms connected to this transformer with all the external services such as air conditioning, IT, lighting and heating connected to the normal power supply system. This will effectively give us a ‘clean’ and ‘dirty’ power supply. It will also help us with our filter design by reducing the induced EMI from our equipment

 

At the time of writing, the filter design has yet to be finalised but our options include:

 

·   Re-use the filters from Aquila

 

·   Buy new filters from a specialist filter manufacturer

 

·   Design a bespoke filter.

 

The first option would be the preferred option in terms of time, cost and technical understanding, the only problem would be the quality of the electrolytic capacitors which may need to be replaced as they degrade over time.

 

Location of and access to facilities

It’s not easy finding a location for Anechoic chambers when you consider the physical size, weight and particular health and safety consideration. The primary drivers for the size and shape of an enclosure is its frequency range and standards it will be qualified against. For TEMPEST and UK Military EMC, an antenna separation of 1m is required with no height scanning. However, for Civil EMC the separation is 3m with the requirement for height scanning the antenna. So, the physical size is set by the Civil EMC requirements and the relative dimensions of the anechoic chamber is determined by the frequency range. Thus, you end up with a very large room! Luckily, we had just the building to accommodate facilities of this size (see fig 3).

 

 

Figure 3: Aerial view of QinetiQ Farnborough

 

Conclusion

Relocating any large-scale facilities is a difficult and challenging feat of engineering and project management, QinetiQ is however lucky enough to have those skills and experience in abundance. However, at times the process has been frustrating but also incredibly rewarding when you are presented with problems and can find effective solutions.  They key to success is a strong integrated team of experts, contractors and managers making full use of previous experience to identify and reduce technical and financial risks.

 

By the time you read this we will be well on the way to completing the relocation and new build and we will be starting the process to meet the requirements set by UKAS and CESG (for TEMPEST). We also very much look forward to continuing to deliver an excellent service to our customers and to do so more efficiently in our new leaner streamlined shape.

 

For more information contact EM Research & Assessment Services on +44 (0)1252 39 2491 or EMCfocus@QinetiQ.com