AIR TRAFFIC CONTROL
The New Scottish Centre
51. The CAA's preferred bidder for the New Scottish
Centre is SkySolutions, a consortium of Bovis and Lockheed Martin,
who had competed with AyrTec, a consortium of Hughes and Laing.
The contract with them has not been placed yet. 85% of the systems
at Swanwick and Prestwick will be common to both.
Despite the problems at Swanwick, NATS believed that the system
being developed for the NERC was a good one and would do what
was wanted of it.
It had stayed with the same system for the NSC because there was
no alternative that would meet NATS' requirements.
52. NATS planned to sign a letter of intent with
SkySolutions very soon after it gave evidence to us on 11 February
and would pay the company an initial small sum of not more than
£5m. It then intended to sign a binding memorandum in April
1998 and the final contract in August once SkySolutions had organised
its financing. NATS assured us that it would not sign any binding
agreement with SkySolutions until it was sure that the software
at NERC could be made to work.
This would be known by April.
53. NATS recognised that there were possible disadvantages
to having only one supplier of software systems and told us that
this was one of the reasons why the contract negotiations had
taken so long. It was anxious to ensure that it would be able
to pay a competitive rate for the several upgrades of the systems
at NERC and the NSC that would be required over the lives of those
was also trying to ensure that the contract gave it some right
to obtain a share of the commercial benefits of the system if
it were sold to other countries.
54. Sir Ronald Mason, former Chief Scientific Advisor
at the Ministry of Defence, was concerned that using the same
contractor for both centres increased the chance of common fault
failure, and that use of different contractors with dissimilar
software, as was done in the case of some commercial aircraft,
could reduce this risk.
He told us that the selection of the same software supplier for
the New Scottish Centre as for the New En Route Centre
"Is inconsistent with the basic principle
of safety-critical software development practice ... Isolating
the Scottish Centre from the NERC project is now essential
to ensure that both systems do not share common software problems
and errors-both in logic and execution-that may only become apparent
when the system has been in operation for some time."
Sir Ronald explained that NERC and the NSC needed
to be interoperable, but they did not need to be identical.
He recommended that an independent assessment be undertaken, perhaps
by the National Audit Office, with the remit to determine whether
NERC met the original specifications and was working reliably
and safely enough to be accepted.
This assessment could be run in parallel with NATS' acceptance
tests. He accepted
that this might delay the placing of a contract for the New Scottish
Centre for about three months.
55. We received evidence from the co-ordinator of
the joint Department of Trade and Industry/Engineering and Physical
Science Research Council Safety Critical Systems Research Programme,
S R Price, that although in an ideal world dissimilar systems
might be an answer, the UK's resources were limited and it would
be more risky to develop parallel dissimilar systems than to do
one system well. He did not think that the strategic situation
of two ATC centres providing back-up for each other was comparable
to that of aircraft on-board systems.
An IPMS witness did not believe that every system in NATS could
realistically have dual software systems.
Professor Ladkin did not agree with the categorical statement
that single sourcing was inconsistent with the basic principles
of safety-critical software development, since safety issues were
more subtle than that. However he did not offer a view as to whether
single- or dual-sourcing was more appropriate for NERC and NSC.
56. NATS' response to Sir Ronald's point was that
it was prudent and safe to have similar, unconnected systems in
the two centres. The chance of both, operating independently,
failing at the same time from the same fault was "so small
as to be acceptable".
NATS believed that an independent assessment of the NERC project
was not justified and would not take two or three months but rather
would delay the project by at least a year by tying up management
The future structure of NATS
57. NATS believed that it would need to invest more
than £100m per year over the next decade in order to maintain
and improve service quality. It is constrained by its External
Financing Limit (EFL) and believes that by 1999/2000 it will have
exceeded its available cash and will be unable to finance its
investment programme other than through the PFI. NATS did not
believe that the PFI was a desirable way to finance its investment
because of the delays caused by its complexity and the risks of
placing a safety-critical service, which relied on an integrated
network of systems, in the hands of a diverse group of contractors.
It also told us that there was little doubt that the use of the
PFI for the NSC would mean that its cost would be "significantly
higher" than if it had been funded through a more conventional
route. This was because of the extra cost of private capital,
the equity element, the premium for the risk borne by the private
sector, the bidding costs (about £10m alone) and the lack
of competition for any upgrading contracts in future.
58. NATS therefore believed that a way of financing
its investment programme needed to be found. It did not think
that increases in its EFL or changes in the definition of the
PSBR were likely and therefore recommended structural change to
the company. It set out three options which would allow adequate
investment funding to be secured:
- A regulated utility:
NATS would be established as a plc and floated, with at least
50% of the shares being sold to avoid problems with the company's
classification by the Treasury. Systems for safety and economic
regulation would be established.
- Trust status: a precedent
was NAV CANADA, now responsible for Canadian ATC. It was sold
by the Government for $1.5 billion CDN, financed by a $3 billion
CDN loan from a consortium of banks. It was required to recover
its costs rather than make a profit.
- 'Corporatisation': precedents were found
in New Zealand, Germany and the Netherlands. The air traffic service
provider would be incorporated as a government company which could
borrow privately and would be expected to behave as if it were
in the private sector and to return a dividend to its shareholder.
NATS believed that this could only work in the UK if the present
PSBR rules were altered.
The CAA and NATS recommended that NATS be privatised
as a regulated utility. In 1995 the then Transport Committee recommended
consideration of the option of corporatisation.
The Government is presently reviewing the future of NATS and will
consult on its conclusions; we were told that the Secretary of
State was "concerned to ensure that NATS has adequate funding
for future investment".
59. Privatisation of NATS was proposed by the previous
government but was opposed by a number of interested parties including
airlines, who objected to the fact that they would have to pay
for NATS' profits within their ATC charges. Although NATS was
not in the end privatised it was separated more clearly from the
CAA by its establishment as a wholly-owned subsidiary of the CAA
on 1 April 1996. The IPMS suspected that the debate about the
privatisation of NATS in recent years had in fact harmed the Swanwick
project: "the previous chief executive [Mr Derek McLauchlan]
was going to war with the previous Government over privatisation
investment and he took his eye off the ball in developing the
GATCO, however, thought that concern about privatisation was not
likely to have been the cause of the problems with Swanwick.
60. We took evidence from witnesses about the different
methods of restructuring in other countries which have allowed
corporations to borrow money privately for capital investment
without being privatised. These witnesses included the Director
of Air Navigation Services and Aerospace Safety Regulation in
Canada, and a former New Zealand Treasury official.
61. The Canadian Air Navigation System was sold to
NAV CANADA in 1996 for $1.5 billion CDN. It is a not-for-profit
corporation with a Board comprising representatives of users,
government, unions and others. It has borrowed $3 billion CDN
privately (which was 3 times oversubscribed) and the corporation
is allowed to collect user charges on the basis of cost recovery
rather than profit.
Any profits made must be used to pay debt or be spent on the air
traffic service. The restructuring was driven by the need for
an efficient and responsive service and by the realisation that
because of budgetary constraints the former air traffic control
service was likely to be underfunded and unable to afford the
It was "widely considered to have been a success for all
concerned": for the taxpayer, through $1.5 billion CND in
receipts, for the industry, in allowing the system to modernise,
for users, who benefited from more efficient operations, and for
employees, NAV CANADA itself and the government of Canada.
62. New Zealand Air Traffic Control was established
as a State-owned Enterprise (SOE), known as the Airways Corporation,
in 1987. The New Zealand government developed an SOE policy in
the mid-1980s in order to address the problems of a number of
state-owned businesses that were loss-making and providing poor
services. Legislation required the SOE to be operated as if it
were privately-owned, and its borrowing was no longer counted
as public sector borrowing. Many of the original SOEs in New Zealand
have in fact been privatised, although the Airways Corporation
remains entirely government-owned.
This is probably because of concerns on the part of airlines about
it being a natural monopoly.
One special feature of the new structure is the official recognition
of the policy tension involved in trying to implement the objective
of commercial autonomy and accountability when the government
is the owner of the company. The requirement to prepare a 'Statement
of Corporate Intent' was included in the SOE legislation in order
to enable a compromise to be reached between conflicting demands
in the event that a minister wished to pressurise an SOE into
pursuing uncommercial objectives.
63. Until 1993, Dutch air traffic control was managed
by and responsible to the Dutch Ministry of Transport. It is now
a free-standing corporation which reports to government. Its budget
is still approved by the relevant minister, but any expenditure
by the air traffic control organisation is not now considered
to be part of government expenditure. This is very close to privatisation
but because the organisation does not have shareholders or make
a profit, its incorporation is different. The minister appoints
the board of management and the advisors. They include representatives
from the Ministry of Transport, Schiphol Airport, the Dutch military,
and the airlines KLM and Transavia. Foreign airlines are not represented.
64. Airline representatives were concerned that reliance
on the PFI might lead to higher costs in the longer run and the
risk of monopoly capture of an important piece of infrastructure.
They stressed that the most important requirement as far as the
structure of NATS was concerned was to free it from the PSBR and
allow it to borrow privately to fund investment.
65. The Minister for Transport told us that the Government
was looking at alternative models for the structure of NATS, although
he did not see any urgency about this. The focus of the Government's
examination would be on investment and whether there was a different
model that would "enable them to achieve the highest levels
of investment in a more sustained and reliable way in the long
66. NATS is part of the CAA whose Safety Regulation
Group (SRG) is responsible for the safety regulatory oversight
of all UK air traffic service providers, including NATS. The SRG
achieves this by considering individual air traffic control units,
their staffing, equipment, maintenance and any other necessary
arrangements. When satisfied that a provider has demonstrated
that the service to aircraft will be safe, the SRG grants a formal
approval under the Air Navigation Order.
Personnel sometimes move between these two parts of the CAA.
67. Some witnesses supported a clearer separation
between the SRG and NATS. These included the IPMS, who thought
that such a separation would be required by European developments
in the near future anyway,
and the Aviation Study Group, which considered that a number of
safety-related incidents which occurred between 1995 and 1997
illustrated "the less than satisfactory safety regulation"
of NATS by the SRG and recommended "the transfer of safety
regulation to the Health and Safety Executive in order to create
greater transparency and to prevent potential conflict between
economic regulation, air traffic services provision and air traffic
services safety regulation".
68. The Minister for Transport said that he was "in
favour of separating regulation from service provision",
but did not see any benefit from moving the SRG to the Health
and Safety Executive.
When giving evidence on railways the Deputy Prime Minister confirmed
that he believed it was generally a sound principle to separate
operational and safety responsibilities.
69. The CAA agreed that moving the SRG to the HSE
would not be helpful and indeed, since it would distance the SRG
from the other civil aviation functions with whom a close working
relationship was vital to effective safety regulation, would have
a detrimental effect.
Witnesses from the SRG also did not think that it would be a very
good idea to take safety regulation out of the CAA because the
HSE operated under different laws and aviation was governed by
a large number of international aviation agreements.
On the other hand, the SRG was not comfortable with the combination
of service provision and safety regulation in one body, because
non-NATS air traffic service providers believed, wrongly in its
view, that the SRG gave NATS preferential treatment. It recommended
that the CAA be separated from NATS.
70. Other countries have separated ATC operations
from safety regulation. In Canada, the restructuring of air navigation
services allowed the government "to end its operational role
while continuing to serve the public interest as the safety regulator.
No longer is there a conflict between an organisation that on
the one hand provides the services and at the same time regulates
the provision of those services".
In New Zealand a new Civil Aviation Authority has been formed,
which regulates only aviation safety, that is separate from the
71. There are two European aviation safety bodies.
The Joint Aviation Authorities, a grouping of European aviation
safety authorities, looks after aviation technical standards.
The other is the very recently formed Air Traffic Management Safety
Regulation Commission. The SRG envisaged that safety regulation
policy would eventually be transferred to those bodies. In the
longer term, it was planned that both bodies would amalgamate
into a European Aviation Safety Authority covering aircraft, air
traffic and airport operations. The UK Government supported this
plan because the proposed body would have the power to make rules
which would be applied in a uniform manner. The present system
required consensus. The SRG expected that the new system would
raise general safety standards in Europe and the UK would make
an effort in the preparatory phase to ensure that that was the
ATC 16. Back
ATC 22A. Back
ATC 20A. Back
QQ 466-7&420. Back
ATC 03, paras 6.1-4. Back
ATC 03C. The DETR told us that "No estimate of the cost
of the NSC using government funding has been made in recent years.
As the NSC PFI project is financially free-standing, no comparison
with a conventionally funded alternative was necessary under Treasury
guidelines." (ATC 01A) Back
ATC 03, para 6.8. Back
Second Report from the Transport Committee, Privatisation of
National Air Traffic Services, HC (1994-95) 36, paragraph
ATC 01, paras 22-3. Back
ATC 19. Back
ATC 25. Back
ATC 25. Back
Information from visit. Back
ATC 28B. Back
ATC 13. Back
ATC 01A. Back
HC (1997-98) 286, Q873. Back
ATC 28. Back
ATC 25. Back
QQ569-71; Department of Transport Report 1997, para 4.11. Back