Case study: “Brandenburg Gate” –

How not to build an airport

Peter Jumpertz, Theron Advisory Group

This document and no part of it may be distributed, quoted, or reproduced without the prior written approval of THERON Management Advisors AG

Preface

The “Brandenburg Gate Saga”, the story of Berlin’s new BER Airport, is not a typical case study of a privately owned and operated infrastructure asset. In fact, it never made it that far. Private investors who had initially demonstrated their clear interest in a public-private-partnership and their capabilities to erect and operate the German capital’s new airport pulled out as soon as they understood the detrimental dynamics of this project. Nonetheless, it is worthwhile to analyze what went wrong, and to take away valuable lessons learned. These lessons will apply mainly to infrastructure assets under construction. In addition, they also shed some light on infrastructure businesses which have been operated for many years but could face major overhauling or even rebuilding – whether planned on purpose or caused by unexpected external forces. Berlin has a long history when it comes to airports. The city’s first airport, Johannisthal, opened in 1909 and had been in operation almost continuously, serving mostly small-sized private aircraft, until it was closed in 1995. Berlin’s main pre-war airport was Tempelhof, which opened in 1923. Located in the very heart of the city, Tempelhof had become famous as the touch point for western allied forces’ cold-war “Operation Vittles”. Tempelhof was closed in 2008, expecting a quick replacement by BER, Berlin’s new capital city airport. After the Second World War, officials of communist German Democratic Republic rushed to open Schönefeld Airport, located at the south-eastern suburbs of Berlin. Schönefeld is still being operated today as add-on capacity mainly for low-cost carrier service. The western city’s new post-war airport Tegel opened in 1948, and still has to carry by far the main load of Berlin’s commercial air traffic of about 27 million passengers per annum.

How not to build an airport

Page 2/12

1 Good idea, bad execution When first sketched out after German reunification in 1991, the master plan for Berlin’s new capital airport was plain and simple. The airport should serve about 22 million passengers p.a. with a peak capacity of 6,000 passengers per hour. The cost of erection was not to exceed €1 bn. In order to avoid prohibitive risk for the taxpayer, it was envisioned from the start to privatize construction as well as operation. Opening was scheduled for the Millennium. None of these goals have been achieved to date. A long row of almost unfiltered change requests have led to frequent delays of planned opening dates, massive cost overruns, and finally, a virtually fully rebuild of the airport’s main terminal. At the time of preparation of this case study, according to the most recent official announcement the opening will take place sometime during the year 2017 and the final cost estimation amounts to €6.7 bn. The author would like to point out in this context that he doubts whether these goals can actually be achieved, and will rather stick to those estimations shown in the figure below, indicating an opening in 2018 with a total cost of €8 bn.

Brandenburg Gate Saga

Year

1991

Schönefeld Airport operating license

Budget (EUR bn) / Planned opening*

enables 0.8 / 2000

“Baufeld Ost” scandal Master plan I 1994

Zoning plan recommendation: Sperenberg

1996

Failed Berlin-Brandenburg merger

1999

Joint dev’t planning Berlin & Brandenburg states

causes Ignored! facilitates

enables

“Consensus decision” location Schönefeld

Zoning plan application

0.8 / 2004

causes

2.8 / 2009

Straight flight path plan Failure to privatize airport 2003

Federal “Traffic Route Acceleration” act

causes “Self-management” decision

2006 Start of construction

Nighttime ban

enables 1.7 / 2010 Zoning plan approval

Noise proofing

causes 2007

Straight flight path plan

1.9 / 2010 causes

2009 “Dry runs” 2012 2013

Postponement of opening

Berlin Mayor & Head of BBG. State taking over Major plan revision Bankruptcy of main planning office * Source: BERtrug.de, ** Recent press publications

How not to build an airport

2.4 / 2011 4.9 / 2013

Current status (12/2014): No opening date, no final cost estimation, no noise protection, insufficient nighttime ban, no flight path plan, airport will never make money!

8.0 / 2018**

Page 3/12

2 Root causes of failure This outcome is a disaster in an economic sense as well as a substantial damage to Germany’s reputation as a country of efficient and industrious people. Observing the events from hindsight and bluntly put, three factors have caused this outcome: • Incapability at key management positions • A true conspiracy of optimists • Prohibitive complexity produced by purposeful decision.

2.1 Incapable management Top executives as well as their supervisors have failed to produce a plan which reflects appropriate assumptions about uncertain aspects of the airport’s business model. The two most prominent examples here are demand forecast and aircraft design: • Demand volumes. All estimations regarding Berlin’s airport passenger demand were way too low. Berlin’s attractiveness after German reunification, its image as a “hip” city, causing a virtual explosion of tourism, had not been taken into account. Initially, BER had been planned to handle 22 million passengers per annum. Tegel Airport and Schönefeld Airport today are serving 27 million passengers. Even before kick-off of construction, it had been quite obvious that the airport would not be able to fulfill this demand. However, the point of no return had already been passed long before the master plan was finally approved – a typical phenomenon of large-scale publicly funded projects. A radical change of plans regarding key design elements is considered a failure in public, and in a political arena face-saving is valued higher than economic survival. • New aircraft design. The floor plan of Tegel Airport from the late 1970s was designed along a hub-and-spoke pattern. This allowed for very short passenger walkways and a one-level layout, maintaining ground level space for luggage handling, installation of technical equipment, etc. More recent airport business models focus on shop revenues, boutiques, food and drinks, and so forth. Consequently, these business models require a shopping mall layout. Separating arrivals from departures is most effectively achieved by two-level passenger logistics. Luggage handling, air flow systems, etc. must be located underground or on roof tops. BER, like Munich Airport, opted for a “shopping mall with aircraft access”. In a shortsighted attempt to confine costs, however, in contrast to Munich, only two main floor levels were planned. A new dynamics, the Airbus 380, comes into play. Catering for such a “flagship” aircraft had been considered a clear priority right from the first announcement of the Airbus Company.

How not to build an airport

Page 4/12

Thus, the main terminal building had to be completely redesigned to accommodate the Airbus 380 which required virtually a rebuild in order to warrant an orderly passenger flow. Making decisions is about saying “no” when facing costly alternatives! Political leaders, however, could not imagine saying no. Instead, they preferred pushing through without careful consideration of the consequences and potential alternatives.

2.2 A true conspiracy of optimists Virtually nobody in a leadership position at BER was in charge of pessimism. • For example, engineering companies tasked to design and build the airport’s huge and highly innovative integrated air ventilation and fire protection system conspired with architects and technical design offices hired to plan and manage the airport’s facilities. They convinced each other that unresolved questions about the complexity of airflow control interaction with smoke disposal could be solved “on the fly” during installation of equipment. In principle, “on-the-fly” engineering is not a bad approach at all – as long as optionality is explicitly managed; however, in this case, the initial plan was totally unfeasible. • Moreover, airport designers and architects conspired with passenger logistics planners. When blueprinting floor plans and passenger flow control equipment, they convinced each other that a computer algorithm would be able to deal with foreseeable but yet unspecified complications from peak load requirements in on-site passenger logistics and baggage ground transportation. From a cybernetics point of view, such an approach would be called “over-optimistic” at best. • Finally, airport management and potential shop owners convinced each other of the scalability of the overall airport layout. They assumed that plans could be changed at minimal cost in case demand would grow beyond existing forecasts. Small and medium enterprise (SME) shop owners were deceived by fortunes to be made, and made it to the short list. Unfortunately, it seems as though most of the SMEs that invested in the airport’s shopping malls are now facing serious trouble, if not bankruptcy.

2.3 Prohibitive complexity produced by purposeful decision The most disturbing missteps, however, were made repeatedly: • Site location Cooperation among the two top layers of Germany’s three-tier government system, including federal, state, and municipal bodies, involved in airport-related planning, approval, and administration issues, is a major ingredient in BER’s recipe for disaster. Federal government wanted to ensure that the international hub function of Munich Airport would not be jeopardized. As a result, a nighttime ban was considered crucial to BER’s existence.

How not to build an airport

Page 5/12

The State of Berlin wanted to keep the airport close to the city limits in order to benefit from business that would flourish around the airport. Driven by its ambition to control the process and ensure proximity to city limits, Berlin’s administration had acquired a large piece of land next to the city’s Schönefeld Airport in 1991 already – years before the site selection would be finally approved! Another location, Sperenberg in the state of Brandenburg, had been favored by all expert opinions mandated during the process. Sperenberg would not pose any risk to 24/7 operation, allow for highly flexible flight path planning and would not create additional cost of noise protection. Most importantly, Sperenberg would have been easy to expand by a third and fourth runway, if required. However, Sperenberg was eventually dropped and the worst possible location, Schönefeld, was selected. Schönefeld’s disadvantages over all other short-listed locations were obvious from the start: – Swampy underground, creating potential for known unknowns and unknown unknowns – Extreme limitations to take-off and approach flight paths as well as to operating hours – Extreme cost of noise proofing and protection – Highest cost of land to be purchased – The epitome of ignorance: the full loss of Schönefeld Airport as a true option • Design Priority was given publicly to architecture and design etiquette which would reflect Berlin’s ambition as a modern capital of a magnificent nation. This meant, for example, that the roof of the airport’s main terminal should not be obstructed by any ugly piece of technical equipment – no exhaust pipes, no compressors, no escalator housings, etc. As a consequence, virtually all technical equipment, which accounts for roughly half of a typical airport’s erection cost, had to be either hidden behind walls or put underground. The imminent cost of this decision: beyond giving up external options planners have now lost all efficient internal options for change requests. In an environment where tough decisions will not be made, change requests take over control. The end result is that this has become a cost multiplier for the project.

How not to build an airport

Page 6/12

3 Lessons learned from an infrastructure asset management perspective 3.1 Sourcing the appropriate human resources Like any other investment, an infrastructure asset’s value gains, regardless of its context, are driven by three steps: • Early and comprehensive identification of critical issues to be resolved • True management decisions under uncertainty • Effective execution of critical decisions. The key factor of success in each of these steps is purely having people with the appropriate skill sets on board. This issue can be regarded as a sourcing problem to some extent. That is how to define what is needed, how to gain access to high potentials, how to select the best of them. BER stakeholders have miserably failed when it comes to managing critical skills and human resources. As the public had to find out step by step, a number of critical HR activities got out of control. In 2014, the Chief Technical Executive at that time had become the main figure in a wideranging corruption scandal. Moreover, the chief technical engineer in charge of designing the highly complex air conditioning system had to admit that he had misled his employer as to his credentials and academic background. Instead of a certified engineer a technical draftsman had designed one of the most critical components of the endeavour. As we know by now, the result is disastrous. Especially after a catastrophic fire at Dusseldorf Airport, which left 17 people dead, mostly from inhaling smoke, German fire regulation has become a real challenge to large construction projects, in particular, airports. BER management was certainly aware of this requirement. And still, they had not been able to deliver on their core responsibility – managing critical issues! One has not to be an expert to manage experts. Commercial management as well as project management are truly general tasks, not expert tasks. The key point here is: What seems to be selfexplanatory for hardware-type assets and natural resources is also fully valid for human resources. There are experts who specialize on sourcing scarce resources. Managers must leverage these expert capabilities. Thus, especially at the beginning of a potential infrastructure investment story, selecting and working with a top talent headhunter is one of the predominant tasks of an asset manager.

3.2 Optimistic pessimism Beyond the fact that managers often hate to deal with worst case scenarios, there are two crucial aspects in risk management. First, in hindsight, worst case scenarios are the most probable reality. Misled by the all-present bell-shaped risk curve, we tend to believe, the worst case is the exception, not the rule.

How not to build an airport

Page 7/12

Second, when developing scenarios and looking for the worst case, analysts typically limit their scope to “bad” and fully disregard “much better than expected”. This bias is also caused by bellshaped mindsets. However, even if the probability of “much better” can be relatively low, the expected value can be devastating. Infrastructure businesses are often platform-type businesses with rather rigid capacity constraints. Platform means that users of the platform’s services share an eco-system benefit. In essence, an individual user’s net benefits will increase with every additional user, driven either by shared cost advantages or by the network effect of user cooperation.1 This interesting property can turn into a substantial risk in combination with capacity constraints, as it will create an immanent propensity to congest.2 In this case, the pressure on a platform’s services will rapidly increase, and users flog to other platforms or render the service useless at all. If you own an asset with such properties, make sure you have options, for example, in the form of scalability or access to neighbouring platforms. This is the only way out of a potential nightmare! This way out had been locked early on at BER because a drastic explosion of airline passenger demand was just not part of the equation. This led to a fatal consequence, beyond poor layout of the new airport. It caused an early conventional-wisdom-type of decision to decommission and close down all other large-scale airports of Berlin.

1

See Wikipedia “In economics and business, a network effect (also called network externality or demand-side economies of scale) is the effect that one user of a good or service has on the value of that product to other people. When a network effect is present, the value of a product or service is dependent on the number of others using it. The classic example is the telephone. The more people who own telephones, the more valuable the telephone is to each owner. This creates a positive externality because a user may purchase a telephone without intending to create value for other users, but does so in any case.” 2

A particularly devilish form of platform gridlock is a context-component constraint in combination with a positive network effect. For example, Parc Astérix of Paris was laid out and built for 4 million customers p.a. After two years of operation, visitor numbers leveled off at about 1.4 million though. What had gone awry? Planners had ignored the gatekeeper function of the amusement parks automated barriers to its main car parking lot. Only two lanes had been provided for, leading directly from one of Paris’ busiest highways into the park. When visitors arrived there in the morning, long bumper-to-bumper traffic jams built backwards onto the highway, letting drivers quickly decide to pass by and go for Disneyland, Parc LeBourget, etc. instead. At the same time, the park itself stayed rather empty. What can be an advantage to a highway – namely vacancy – can become a real deal breaker in an amusement platform. Loners in dark rides have no fun! So, a negative bandwagon effect had been kicked off, and the park slid rapidly towards bankruptcy. Park management was convinced poor attractiveness was the root cause, and tried to raise more capital for new rides. Would they have put themselves in the shoes of their potential customers – as a young management consultant did, when approaching the park riding with his family in a French “Rent-a-wrack” car – they would have saved money, time, as well as their jobs. (This could be a sidebar case study).

How not to build an airport

Page 8/12

The point of no return of the decommissioning of Tempelhof Airport had been reached at some point between 1992 and 1993. The public had been informed that this valuable piece of land in the heart of Berlin would be vacant soon, and that new and highly attractive uses would be developed soon. All citizens were informed they would benefit. Today, Tempelhof remains an open park that is used for flea markets and outdoor concerts. The point of no return of Tegel Airport had been reached substantially later, but still too early to maintain the unit as a valuable option. Even today, after the quagmire is obvious, no politician is able or willing to reverse the decision, as most Tegel community stakeholders have invested both financially as well as emotionally in the airport’s final closure. Real estate companies, for example, have invested in planning efforts, have purchased land lots, buildings, even business in order to participate in the local housing boom which is to be expected after decommissioning of the airport. Other businesses have adjusted their development plans for new facilities and factories in Berlin, hoping to capture a vacant piece of Tegel at bearable cost. Rolling back these decisions would initiate a wave of damage claims against the City of Berlin. So, by ignoring the “best business – worst case” reality, a gridlock has been set up which is costing dearly. How can an asset manager avoid a gridlock? In fact, it is quite simple once you are aware of this type of risk and as long as you understand sensitivity analysis. Build your scenarios, and as soon as you feel comfortable with scenario design, reverse each parameter’s ratios one by one, in essence, by multiplying ratios by -1. Then check whether you still feel comfortable. Take BER: 22 million passengers p.a. had been forecast in the expected case. According to one of the planners involved, one of the so-called “worst cases” showed close to 17.5 million passengers, which is a ratio of roughly -20.5%. The airport would lose money. Plans were made to raise demand in this case. The best case went up to somewhere near 24 million passengers. Profits would skyrocket. Capacity would be snug but even during peak load all key platform components would still be able to operate as planned. Had planners used a “constructed worst case” of 22 million passengers times 120.5%, giving 26.5 million passengers p.a., they would have felt radically constrained. One might argue since nobody would have taken this number serious nobody would have thought about the consequences either. But this is not necessarily true as long as you take a pessimist’s point of view while posing as an optimist. The pure nature of real optimists on-site will take care of the problem. A pessimist who argues with optimistic assumptions earns enormous credibility among optimists. His voice will be heard, and optimism soon will take over again – this time for the good of creating and maintaining valuable options.

How not to build an airport

Page 9/12

3.3 Simple – Simply simpler Obviously, an airport is a complex matter. In general, infrastructure is a complex issue. And since infrastructure is tangible, which means it hurts when you kick against it, you cannot change anything about, can you? Wrong! Simplification is a matter of will and skill rather than feasibility. Simplification is all about tough management decisions, not about immobile facts. If asset managers focus on value instead of hardware, they can find alternatives for management decisions. What this takes beyond analytical skills is the will to take risks. The nature of decisions is in their opportunity cost. A costless decision is of no value as well! At BER, a number of complexity reduction decisions can be identified in hindsight. Starting with the really simple ones, even the difficult ones look easy compared to typical business decisions under uncertainty. • Building design: avoid bottlenecks and poor flexibility arising from “clean and cool design etiquette”, and put technical equipment where it belongs in an airport building! The opportunity cost is close to zero, since passenger demand, shop owner sales, and airline’s willingness to pay service fees – the main revenue drivers of an airport – will not depend on design. • Systems integration: ensure functionality by fully separating air conditioning systems, automated passenger emergency control, smoke disposal, and fire extinguishing equipment both physically as well as with regard to operational control! Even in hindsight, the advantage of an integrated system is not clear. Integration always comes at substantial complexity cost. It has long been known that a complexity cost curve typically has a progressive shape, quickly leading to prohibitive effort in design as well as in operation. Whatever opportunity cost would occur from a simplification decision, they are unlikely to be above its benefits. • Aircraft connectors: ignore two-level aircraft like Airbus 380! What will be the consequence? If Lufthansa will decide to blank out Berlin from its route network, passenger volume will drop from 27 million to about 26.93 million p.a. – resulting in less of peak load problem, less of noise proofing problems, less of nighttime ban issues, shorter take-off slots from reduced vortex, cheaper runway constructions and maintenance, standard refuelling procedures, etc. Such a decision to ignore Berlin will give Lufthansa a headache, as well as all other airlines willing to make money with serving Germany’s capital. Again, quite an easy decision from BER’s point of view with almost no cost but huge benefits associated. • Site location: select Sperenberg instead of Schönefeld! The benefits of such a decision are obvious. But what is the opportunity cost? If one believes in the value of artificial monopolies – protection of the hub business of Frankfurt/Main Airport and Munich Airport – one might call this value opportunity cost. However, protecting these businesses could also have been achieved simply by regulating BER’s operating license. It has already been done at other large-scale airports.

How not to build an airport

Page 10/12

If the City of Berlin as a major stakeholder wants to capture a substantial piece of the new airport’s economic contribution – beyond its principal role in the city’s development – this could have been achieved by creating a side business for Berlin, for example by introducing a highway toll on the newly built airport highway. Again, a decision with relatively low opportunity cost. Simplification can become a key value driver in any kind of non-price-regulated infrastructure assets. The reason lies in the nature of complexity itself. Complexity will reduce transparency for the average user, because it is rather impossible to understand causal relationships between cost drivers on the one hand and individual value on the other hand. Above a certain margin of complexity this will lead to a rapidly decreasing propensity to pay caused by a growing sense of being treated in an unfair way while not being able to avoid freeloaders.

4 Enemy mine From an infrastructure investor’s or asset manager’s perspective, it can be worthwhile to point to a well-known but probably not well-understood fact of life: politicians are not managers. Managers must create transparency and make decisions in a given setting, based on costs and benefits. Politicians must understand key forces at work and shift power in order to achieve their declared targets. Cost in political terms has nothing in common with a balance sheet, but is purely reflected in the balance of power. Losing voters is the real loss on a political bottom line. This typically leads to a simple consequence. If a project goes south, the politician has to start the blame game in public. In order to win this game, the smart public leader has prepared his battle field as cautiously and foresighted as a chess player his end game. So, whatever his role and his position, a politician is a manager’s enemy – always – no exception to this rule! This fact of life brings about two reasonable results: 1. Putting politicians in charge of complex business endeavours is a recipe for economic disaster. While for managers a complex endeavour poses high economic risk, politicians involved will definitely run it into the ground. If you find politicians on boards of infrastructure asset projects exposed to economic risk, keep off! 2. Having politicians on boards of regulated infrastructure assets is an effective money machine though. When the going gets tough, politicians will tend to ease their enemies’ minds by simply buying their support. As they will pay with public money, they will always deem this route the cheapest. Now one might ask the fair question: Why, then, did Berlin’s former Mayor not succeed in saving his position after BER’s problems surfaced? The answer supports our reasoning above: It was not burned public money which finally fired Klaus Wowereit’s ejection seat. Instead, he had to pay with Germany’s reputation as a global leader in technology development. This currency is ex-

How not to build an airport

Page 11/12

changed at par for pure political prowess – by far too valuable to be ignored by his peers and superiors across all parliaments and parties.

How not to build an airport

Page 12/12