Memorandum by National Air Traffic Services
Ltd (NATS 02A)
TRANSPORT SUB COMMITTEE INQUIRY INTO NATS
Thank you for your letter of 9 November in which
you confirmed the additional information promised in the course
of the hearing on 8 November and the additional questions on which
you now require responses.
A list of the "bugs" which remain in
the system at Swanwick
The attached annex provides a list of outstanding
Programme Trouble Reports (PTRs) as at 15 November. The number
of PTRs required to be resolved prior to Technical Handover next
month has continued to fall, from the 120 reported by Colin Chisholm
at the hearing on 8 November to 99 now. The project remains on
track to reduce this to a manageable number of around 30, all
of which will be in a "known state" at Technical Handover.
Details of the shortage of engineers in NATs,
particularly at LATCC
The budget for engineers at LATCC for October
2000 was 215 against 210 staff actually in post. LATCC is currently
recruiting to fill 11 posts which cover these vacancies and forthcoming
retirements. Watchkeeping staff numbers are at full strength.
As far as the situation in NATS Software Services is concerned,
the position is that for the same period there were 204 staff
in post against a budget of 207. In overall terms, there is no
shortage of engineers in NATS.
The computer failures at LATCC this year have,
however, resulted in some diversion of resources in the NATS Software
Services area in order to resolve NAS software engineering issues.
These staff, with their in-depth knowledge of the NAS system,
represent a scarce, expert resource. Their diversion has created
scheduling issues for some other planned tasks, but the team has
just delivered the latest version of NAS software into operational
service only one week behind schedule.
The manpower situation at LATCC, particularly
during the conversion period as controllers train at Swanwick
For Area Control operations at LATCC, NATS currently
has 360 air traffic controllers available to meet a budgeted requirement
for 300 controllers to run the operation at full strength. The
overbearing of 60 controllers has been in place for some time
in order to meet NERC development requirements leading to the
different concept of operations at Swanwick. Next year the overbearing
will be used to provide cover for operational conversion training.
By June 2001, the actual number of controllers is anticipated
to increase to 374 against the same complement of 300 posts required
for Area Control operations at LATCC.
For Terminal Control operations, NATS has a
full complement of 255 controllers to meet current operational
requirements, although there will be a slight shortage (10 to
14 people) after the introduction of two new sectors in the Spring.
NATS managers are confident that this is manageable through the
Summer of 2001.
During this Winter, there will be a number of
staff working off-site on training activities and on the evaluation
of these planned re-sectorisations. In addition, the Swanwick
operational conversion training will start in January. Throughout
this period, the Area Control and Terminal Control operations
will be manned to meet seasonal requirements. As explained in
NATS' memorandum to the Committee, there is a very detailed programme
in place to meet conversion training requirements. A roster for
all LATCC Area Control staff has been published, including over
21,000 individual allocations of staff to activities.
How many near-misses or loss of separation incidents
have occurred during periods of computer failure this year? How
many commercial aircraft have opted to fly in uncontrolled airspace
during computer failures at LATCC to avoid delays?
No Aircraft Proximity (Airprox) incidents have
occurred as a result of NATS computer failures this year. The
failure of the NAS system on 9 June has, however, been cited as
a contributory factor in an incident on that day involving the
loss of standard separation between two aircraft in the Dover
We have no data on the number of commercial
air transport aircraft opting to fly in uncontrolled airspace
during periods of computer failure. However, the number could
be expected to be small, given the fact that nearly all major
UK airports lie within controlled airspace.
The internal costs to date of the Swanwick Project
There is no disparity between the NATS and DETR
memoranda in respect of the costs of the NERC project. The Arthur
D Little (ADL) report explained that the approved budget for NERC
was £474.9 millionmade up of £350.7 million capital
and £124.2 million revenue.
As noted in both memoranda, ADL found that the
expected costs to completion of the project would amount to around
£625 million, ie £150 million more than the original
budget. The ADL estimate of £625 million included approximately
£60 million of site revenue costs, which NATS considered
should not have been included in the total. These costs relate
to expenditure with third parties on the management of the Swanwick
site over the period from the original 1996 "O" date
until the Winter of 2001-02 and hence are unrelated to computer
system development costs. The figure of £625 million has,
however, come to be generally regarded as the revised target figure
for the project.
ADL also identified an additional £76 to
£86 million of internal NATS staff costs associated with
the management of the project and the operation of the Swanwick
site. ADL did not include these internal staff costs within the
project costs identified above on the grounds they had always
been accounted for separately from the project. It is these costs
that are referred to as "NATS internal costs" in DETR's
The Tender Evaluation Board which dealt with the
New Scottish Centre contract
This question is assumed to relate to the tender
evaluation process which took place to determine a preferred bidder
for the New Scottish Centre (NSC) Private Finance Initiative (PFI)
contract in late 1996/early 1997. The PFI contract has subsequently
been abandoned and work is now proceeding on the basis of revised
funding and procurement arrangements.
The original procurement and evaluation processes
for the NSC were carried out in accordance with an invitation
to Participate (ITP) using the Negotiated Procedure under the
Public Works Contract Regulations 1991. In addition, the NSC tender
evaluation process was structured to conform to the Government's
PFI guidelines. Ministers of both the Department of Transport
and HM Treasury endorsed NATS, procurement process both before
an advertisement was placed in the European Union Official Journal
and before the ITP was issued. The ITP stated that the Civil Aviation
Authority intended to award the contract to "the most economically
advantageous offer" received from tenderers taking into account
the evaluation criteria listed in the ITP.
The first stage of tender evaluation, based
purely on technical and commercial considerations, led the Tender
Evaluation Board to conclude in late 1996 that:
"The prices declared by the tenders for
the construction and lifetime support of the New Scottish Centre
make this programme financially the largest that NATS has ever
undertaken. Before entering into such a commitment, NATS must
be confident that the chosen Project Company will be able to fulfil
the technical and commercial requirements. On the basis of the
evaluation, and paying particular attention to the performance
of the systems companies on previous ATC contracts, the Tender
Evaluation Board does not believe that the necessary level of
confidence has yet been established to permit the selection of
the preferred tenderer."
Thus the advice of the Tender Evaluation Board
was that a decision on the preferred bidder should be deferred
sine die pending confirmation on the ability of the competing
bidders to deliver air traffic systems projects then in progress.
Strategies for handling this situation were
considered by management, with one option being to delay the letting
of the NSC contract for up to a year to assess the rate of progress
on the bidders' other projects. However, in parallel with the
tender evaluation process, a review of the NATS Two Centre Strategy
had concluded that the NSC should remain a core component of the
company's Centre Development Plan and that the project needed
to proceed without delay in order to meet operational requirements.
In the light of these conflicting considerations,
NATS management concluded that the project should go ahead if
mitigation of the Tender Evaluation Board's concerns could be
achieved. While evaluation of the technical, commercial and total
lifecycle costs issues between the bids were considered to be
broadly neutral, the key discriminators between the two tenderers
were thus credibility and risk. The selection of the Preferred
Bidder, Sky Solutions, was made in accordance with these criteria.
This selection was endorsed by the CAA Board and subsequently,
in March 1997, by the then Secretary of State for Transport. The
Centre Development Plan was endorsed by the new Government in
The Chairman of the Tender Evaluation Board
at the time has no recollection that he threatened to resign over
this decision and he continues to be employed by the Company in
a senior position.
Why did the PFI for the New Scottish Centre eventually
The PFI contract for the New Scottish Centre
failed for three reasons:
the costs of the proposals were higher
than NATS' airline customers were prepared to pay. In particular,
the PFI structure would have led to a sharp increase in charges
at the time the Centre entered operational service;
it proved impossible to reach agreement
on some elements of the PFI contract that met NATS' operational
and commercial objectives at the same time as meeting the Government's
PFI criteria, especially in relation to the transfer of commercial
risk to the contractor. Despite many meetings with the Treasury
Task Force, DETR officials and senior accounting advisers, the
balance sheet treatment was never finally resolved; and
following the announcement of the
PPP proposals it was accepted by Government that to have a private
sector NATS with one of its main centres in another company's
ownership was not logical.
The main issues however, were affordability
and the appropriateness of using the PFI model for major IT projects.
The structural rigidities of the PFI concept are perhaps well
suited to the provision of fixed infrastructure but not to the
development and maintenance of complex IT systems that require
to be changed and modified many times over their system lifecycle.
The NSC PFI was intended to run for a period of 25 years (development
and operation), resulting in NATS' customers paying for unnecessary
contractor overheads, and with no opportunity for NATS to share
in any cost reduction opportunities later found to be available.
The PFI programme relied on the continued co-operation of a single
contractor whilst providing little leverage to NATS to control
the activities of that contractor. In particular, the PFI structure
allowed no opportunity to compete future maintenance and development
The importance of competition has been well
illustrated by the re-tendering of the NSC building design and
construction activity. The building costs now seem likely to be
reduced significantly as compared to the PFI project plan. In
addition, system design and development has been broken down into
discrete Contract Line Items (CLINs) subject to competition and
incentivisation arrangements. Under the Master Agreement negotiated
with Lockheed Martin (LM), the entire scope of services that make
up the Systems Programme, is the sum of individual CLINs. Each
of these may be awarded and managed separately. It is possible
to have a mixture of contract forms, ie. cost reimbursable, fixed
price, cost plus incentive fee or any other appropriate contract
type for that element of work.
Overall it is confidently expected that it will
be possible to achieve savings in excess of £100 million
on the overall project costs as a result of changing the procurement
Details of the savings that have resulted from
Lockheed Martin providing the software both for the New En Route
Centre and the New Scottish Centre
The best independent analysis of the position
on this issue was provided in the Arthur D Little Report. ADL
concluded that "the benefits of having a single system supplier
for NERC and NSC far outweigh the drawbacks. Future upgrades can
be implemented on both systems, realising substantial savings,
and support and maintenance costs can be minimised, again with
substantial savings. Also the development risks when using the
same supplier will be much reduced".
The ADL Report makes the following further points:
"There are many benefits to be had by choosing
the same supplier as for NERC, and since the tender evaluation
stage for NSC it has become very apparent that these benefits
are significant. The primary benefits are as follows:
future upgrades can be implemented
on both systems, realising substantial savings for NATS, estimated
by NATS to be more than 30 per cent.;
substantial savings can be made in
maintenance and support costs since only a single engineering
support cell is required. Over the lifetime of NSC (18 years)
this is estimated by NATS to be worth £200 million; and
there will be much higher confidence
that the NSC system will work. The system architecture will be
as for NERC (the CXSS software, application software and hardware
platform will all be common) and therefore the risks will be much
smaller than if a different supplier were chosen."
The ADL Report also noted some drawbacks, relating
in particular to the risk of "lock-in" to a single supplier.
The ADL report noted that "Management of the business risk
is key to ensuring that the drawbacks of having a single system
supplier are minimised and NATS financial exposure is limited.
One important aspect of this is to ensure that the right commercial
arrangements are in place for NSC. Although it is not within the
scope of this study to investigate the merits or otherwise of
the PFI, the mechanism by which it is understood NSC is to be
financed, choosing this way forward for NSC will inevitably impact
on the business risks for NERC.....On the face of it, the potential
for "lock-in" is increased by PFI."
Subsequently, as noted above, the Government
decided that the PFI should be abandoned. The revised procurement
arrangements described above will significantly reduce the potential
for "lock-in" to a single supplier as well as resulting
in further savings beyond those envisaged at the time ADL completed
What costs have been incurred by NATS, the CAA
or any other public body as a result of the decision to opt for
the PFI for the New Scottish Centre, and then to cancel the contract?
In May 2000 NATS settled a claim with Sky Solutions
for work undertaken in respect of the NSC PFI contract. The settlement,
totalling £7 million, was approved by the DETR. No payment
was made to the bidders for work undertaken prior to March 1997
when the preferred bidder was announced.
Prior to the announcement in November 1993 that
the NSC would be funded under the PFI, the project had been proceeding
on the basis that it would be conventionally funded and procured.
Following the announcement, the specification had to be re-written
as an output specification, service level specifications and mechanisms
had to be developed, the arrangements had to be tendered, and
the resulting bids had to be evaluated and negotiated. This activity
added several years of delay to the project and hence significantly
increased the level of project support costs. In addition, the
costs of specialist advisers would need to be added to the cost.
The costs of sustaining projects to support the current operation
are set out below. It is not possible to calculate precisely what
the difference would be between the project cost had it been funded
and procured conventionally, on the basis of the timetable originally
envisaged, and the additional cost arising specifically as a result
of the PFI.
What pressures will be put on the air traffic
control system due to the delay in opening the New Scottish Centre?
What additional costsin terms of replacing and updating
systems in the existing Centrewill be incurred?
As a result of the delays in completing the
New Scottish Centre project it has been necessary to implement
a programme of work at the Scottish Area Control Centre (ScACC)
at Prestwick to sustain operations until the New Scottish Centre
is brought into operation.
This programme of work is aimed at addressing
three areas, namely, ATC capacity, systems capacity and systems
ageing. The key milestones in the programme are:
new bank of workstations constructed
for military occupation, year ending 2000;
new discrete Military Voice Communication
System ready for operational service, March 2001;
Radar Data Processing and Display
System ready for operational service, March 2001;
new sectors to be ready for operational
service winter 2002 and 2003; and
new Domestic Operations Room layout
to be completed end 2003.
The total capital cost of this Sustainability
Programme is £6.4 million plus internal NATS Engineering
costs of £2.3 million. The programme ensures that ScACC will
continue to operate a safe and effective service to NATS' customers
in the period to the transfer of operations to NSC.
What benefits will the New Scottish Centre bring,
in terms of increased capacity in the airspace it covers, or in
terms of reduced delays?
The New Scottish Centre is one of the key elements
of NATS' Two Centre Strategy endorsed by Ministers in May 1997.
The centre will provide the platform for increasing airspace capacity
throughout the life of the building, over the next 30 to 40 years.
The capacity benefits arise from the ability physically to increase
the number of sectors in Scottish airspace throughout this period,
and the ability of modern technology to support the task of controllers
through the development of productivity tools. The current centre
can accommodate a limited number of additional sectors, but there
will then be no further space in the operations room for subsequent
expansion. There are three other major benefits:
existing ageing equipment systems
will be replaced, thus enhancing the overall reliability and resilience
of the system.
commonality with Swanwick will enable
contingency cover to be provided in the event of the catastrophic
failure of either centre; and
common systems and software will
reduce maintenance and upgrade costs, which are expected to become
an increasing factor in NATS' overall cost base.
Why did EDS default on the Oceanic Flight Data
Processing System contract?
EDS failed, in NATS' view, to meet the requirements
for Critical Design Review 2 by 31 May 2000 as mandated in the
contract between NATS and the Company. This assessment is not,
at this time, accepted by EDS.
What cost has been incurred by NATS, the CAA or
any other public body as a result of the decision to opt for the
PFI for the New Oceanic FDPS, and then to cancel the contract?
Costs have been incurred with specialist advisers
and on internal project support costs. NATS has also had to incur
costs on sustaining the current system and has let a number of
support contracts which are currently under review. NATS is currently
in discussion with EDS about the value of any elements of work
completed to date that might be of benefit in developing a new
system, and about contractual issues arising from the Critical
Design Review 2.
What "off the shelf" flight data processing
system do you intend to purchase for the Oceanic centre? What
consideration have you given to purchasing other systems, such
as those to be developed for the New Scottish Centre, and those
at the New En Route Centre, "off the shelf" rather than
having "bespoke" systems designed?
The NATS Board has agreed a strategy of procuring
a modified Commercial Off The Shelf (COTS) product to meet NATS'
core requirements for the Oceanic Flight Data Processing System
with a progressive upgrade of the system to meet future requirements,
It is expected that a competitive detailed project
definition activity will be undertaken with a view to defining
the project in terms of "core" requirements. The objectives
of this Project Definition activity include:
reaching an agreed and validated
set of Requirements for the project which meet NATS' business
needs and are both practicable and affordable;
confirming that one or more COTS
products can, when modified, meet NATS' requirements and be capable
of further modification to meet NATS' requirements; and
gaining assurance that products proposed
by one or more suppliers can meet NATS' requirements for the project
and that subsequent modifications can be delivered, as planned.
The NERC and NSC systems are not suitable for
the Oceanic task in the medium term. The Oceanic task is wholly
procedural requiring advanced conflict prediction and resolution
which is possible in this highly predictable environment. In the
longer term (10 years plus), it may be possible to use a common
system for both Oceanic and domestic operations but this is beyond
the concept of operation and the current generation of systems
being put into operation.
What has caused the "marked improvement in
the reliability of the current system" at the Oceanic centre?
Do you anticipate further improvements in future?
The current system entered operational service
in 1987 but was beset by serious hardware and software reliability
problems. A hardware upgrade programme in the early 1990s improved
matters and subsequent software upgrades have continued to improve
the reliability of the system. We do not anticipate any further
significant improvements in the functionality of the system as
future developments to cater for CNS/ATM (Communication Navigation
Surveillance/Air Traffic Management) in the North Atlantic region
depend on the technological platform afforded by FDPS 2. The current
system operated with 99.93 per cent availability in 1999.
Can you explain more fully the nature of the "consolidation
plus tools" investments plan that you have made your preferred
option? Why is the plan described as "high risk"?
This investment plan option, referred to in
Section 1.3 of the NATS Service and Investment Plan, seeks to
achieve the necessary ATC capacity to meet forecast demand through
the earliest possible introduction of computer assistance tools
to increase ATC sector productivity in both London and Scottish
Area Control. It also aims to achieve the consolidation of all
of NATS' facilities, including London Terminal Control and Manchester,
in the two Area Control Centres at Swanwick and Prestwick by 2010.
This plan therefore delivers the capacity goals whilst addressing
overall operating costs.
The reference to high risk does not refer to
any safety issues. The simultaneous programme to deliver both
the capacity and unit consolidation elements is very challenging
in terms of technical requirements and timescales and is consequently
rated as high risk to maintain the proposed timetable.
24 November 2000