Underground construction projects differ from other construction types and are project-specific due to the environments in which the construction takes place. The underground environment is diverse and fewer risks are foreseeable prior to project implementation. The structure of such an underground work is influenced significantly by nature. Rugged geology co-exists with structures built through human efforts. Therefore, a successful outcome can only exist where there is a mutual symbiosis between natural and human factors.

Underground construction contracts must, therefore, anticipate a higher level of risk which can never be fully eliminated. As the work takes place in natural surroundings, the characteristics and behaviour of these surroundings are not always foreseeable. A final prognosis can only be defined during the realization phase.

Geology-related anomalies are ranked among the major hazards, for example, excavation instability. Should the necessary measures not be taken correctly and in time, these instabilities can spread and reach breaking point. Depending on the particular conditions and size of the rock cover, breaking point can result in cave-in phenomena that pose a threat to surface structures above the underground works.

The largest underground construction works are usually tunnels. The following problems are ranked among the main tunnel construction hazards:

• loss of tunnel face stability;
  
• portal collapse;
  
• collapse of tunnel ceiling (ground arch) at the heading, resulting potentially in:
  
  
  
  • excessive overbreak;
    
  • ceiling collapse up to the surface.
  
  
• face fall-out on tunnel;
  
• low stability of tunnel face;
  
• tunnel bottom growth, lining pervasion into soft subsoil;
  
• excessive growth of convergences – tunnel profile squeezing, primary lining deformations;
  
• excessive ground water inflow into the tunnel;
  
• sudden water/mud/runny sand breakthrough into the tunnel;
  
• dangerous gas or radiation bursts into the tunnel from:
  
  
  
  • methane;
    
  • natural gas from ruptured piping;
    
  • CO₂.
  
  
• occurrence of stray currents;
  
• excessive surface sinking above the tunnel and related impacts on surface structures, power and service utility lines;
  
• drawdown, destruction of water wells around the tunnel;
  
• damage and destruction of water courses near the tunnel by mine water discharges that may have substantially changed their chemistry (for example, concrete extracts);
  
• damage due to pressure grouting compacting the rock massif or due to anchor grouting (damages to power and service utility lines, surface swelling);
  
• improperly selected and implemented tunnel insulation and water infiltration into the tunnel.

The first national document that deals integrally with how to assess, analyse and control the above risks is The Joint Code of Practice for Risk Management of Tunnel Works in the UK issued by the British Tunnelling Society (BTS) in September 2003. A closely related document striving to become internationally renowned is A Code of Practice for Risk Management of Tunnel Works, prepared by the International Tunnelling Insurance Group in January 2006 (the Codes). These Codes are not enforceable or legally binding. However, they do contain some important principles which endeavour to make all parties involved take the right approach to risk identification, control and elimination.

The Codes came into being through the cooperation of the Association of British Insurers, Assurance Companies, BTS, the International Tunnelling Association (ITA) and the International Association of Engineering Insurers (IMIA) also in reaction to some of the larger insured accidents that include (in US$): the Great Belt Link Fire (Denmark, 1994, damage $33 million), Munich Metro Collapse (Germany, 1994, damage $4 million), Metro Taipei Collapses (Taiwan, 1994 and 1995, combined total damage $24 million), Metro Los Angeles Collapse (USA, 1995, $9 million), Hull Yorkshire Collapse (UK, 1999, $55 million), TAV Bologna-Florence Collapse (Italy, 1999, damage $9 million), Anatolia Motorway Earthquake (Turkey, 1999, damage $115 million), Metro Taegu Collapse (South Korea, 2000, damage $24 million), TAV Bologna-Florence Collapse (Italy, 2000, damage $12 million), Taiwan High Speed Railway Collapse (Taiwan, 2002, damage $30 million), SOCATOP Paris Collapse (France, 2002, damage $8 million) and the Shanghai Metro Collapse (China, 2003, damage $60 million).

Particular mention will be made of the collapse which took place during the construction of the Heathrow Express Rail Link. On 21 October 1994, the tunnel at Heathrow Airport in London caved in and became one of the most extraordinary construction events of the last quarter century. The collapse caused the cancellation of hundreds of flights, a six-month delay of track commissioning and caused damage totalling more than $141 million. The court ordered a record fine of £1.2 million against the contractor Balfour Beatty plc for endangering the safety of the public and for gross violation of occupational safety. On the supplier of geo-monitoring (the Austrian company, Geoconsult GmbH), the court imposed a fine of £500,000. In addition, the court ordered each company to pay a further £100,000 in legal costs.

Large, insured accidents make tunnel construction more expensive. The Codes, therefore, aim to unify and determine a minimum standard of risk control methodology. The Codes cover all phases of underground construction efforts, i.e. the preparations, engineering, project allocation and implementation. The Codes emphasize the insurers’ involvement in the contract. An insurance company is authorized to perform site inspections and require a remedy should discrepancies be found.

The following are among the main principles of the Codes:

• The requirement to submit the Register of Risks. The Register of Risks is an open document (it is possible and desirable to extend it during the course of construction), which clearly defines to whom a risk belongs, how it is to be controlled and how it is to be mitigated. The Register of Risks is a part of a Quality Control System, being, as such, subject to independent audits.
  
• Use of the ‘standard forms of contract’ and technical standards.
  
• The contract should include a risk allocation and sharing clause, concerning geology or unforeseeable physical conditions.
  
• The contract should include a provision regulating the geo-monitoring process.
  
• The contract should include a provision allowing variations and implementing value engineering.
  
• The employer must have sufficient knowledge of geological risk control. If lacking this knowledge on the side of its own staff, they are obliged to hire a consultant or a contractor able to meet this requirement.
  
• The employer is obliged to invest sufficient funds in geological and hydro-geological surveying. This allows bidders to prepare and price the offer in respect of the known ground conditions and related risks in connection with tunnelling.
  
• The employer is also obliged to have sufficient funds and time for project preparation.

A project may be deemed ‘uninsurable’ where the above requirements are not met. This, in itself, is a project commencement obstacle.

Underground construction risk tends to be allocated to contracting parties through standard forms of contract. The risks in connection with unforeseeable ground conditions are usually borne by the employer and the technology-related risks by the contractor. It is nevertheless important to know the exact wording of the specific conditions or modifications of the standard form within every individual project.

Mere use of any standard contractual conditions will not eliminate unforeseeable risks and lack of proper project preparation. In general, three approaches to unforeseeable (physical conditions) ground conditions risk allocation can be distinguished and are discussed below:

1. Full risk of ground conditions is borne by the employer. Given such a risk allocation, the actual costs incurred in connection with the work completion under the encountered ground conditions, agreed headquarters, site overhead costs and lost profits are paid to the contractor regardless of total expenses.
   
2. Sharing of ground conditions risks. Given such a risk allocation, the contractor is compensated for all actual costs incurred (regardless of the total) in connection with work completion and agreed headquarters and site overhead costs. The contractor is not, however, compensated for lost profit in connection with adverse ground conditions encountered if they differ from the expectations spelt out in the terms of reference.
   
3. Full risk of ground conditions is borne by the contractor. Given such a risk allocation, the contractor is not compensated for the actual costs incurred in connection with the work completion, agreed headquarters and site overhead costs and lost profit. This approach is not recommended for underground construction projects.

Where the employer does not invest any funds into geological surveying and risk analyses, the bidders must do so at their own expense, resulting in increased costs in connection with submitting the offer and, inherently, restricting the number of bidders. Bidders must be provided with enough time and space to carry out a site inspection and investigation. In the case of underground projects, only a part of the site is available for inspection. The technology employed in construction may prove unsuitable, should the employer assign all the responsibility to the contractor who, in turn, comes across unforeseeable ground conditions during underground works. In such cases, the economic sustainability of the bid may collapse as a whole.

A contractor bearing the risk of unforeseeable ground conditions is forced to adapt the project and method of construction in line with expected economic returns. In the event of a high level of losses, the contractor will almost certainly look for all possible ways to terminate the contract prematurely. This can impact on the total costs, quality, safety and service life of the entire tunnel (see the recent case of a tunnel construction early termination in Obrascon Huarte Lain SA -v- Her Majesty’s Attorney General for Gibraltar [2014] EWHC 1028 (TCC).

It is always difficult to accurately evaluate the nature of ground conditions. In fact, nature of itself is never completely predictable nor are natural events completely foreseeable. This, in itself can give rise to disputes.

The following should be taken into account whenever the foreseeability (or lack of) is to be assessed:

• information generally available at the ‘before invitation to bid’ stage;
  
• information provided by the employer (such as geotechnical surveys);
  
• results of the contractor’s own surveys at the bid preparation phase;
  
• subjective factors, i.e. what a well-experienced ‘reasonable’ contractor should and could foresee, based on the above information, sources and site investigation.

When an issue as to what was or was not foreseeable is to be decided, the most significant argument, however, is what the contractor had allowed or enabled through their own on-site actions and how they had responded to the resulting situation. In respect of early warning principles, the rule is, therefore, to notify the employer of any unforeseeable conditions at once, whenever they appear, and to coordinate the steps to be taken to eliminate the related risk of non-compliance with the contract.

Another useful tool for analysing what is, and what is not, foreseeable may be a requirement placed upon the bidder to submit (along with their bids) all information and data used during bid preparation. This information will then become available to the other bidders. For example, subsequent surveys performed by the individual bidders and details of expenses incurred in connection with them. Should an unsuccessful bidder (at the preparation phase) assess a particular hazard as foreseeable, the successful bidder cannot claim the same hazard as being unforeseeable. These pieces of information can be considered when variations are to be assessed and the validity of the contractor’s claims evaluated. On the employer’s side, however, it is always necessary to take all appropriate measures to prevent abuse of any sensitive commercial information.

The regulation of unforeseeability differs between particular jurisdictions. In international construction projects, FIDIC has laid down a standard for dealing with unforeseeability in contracts. Regarding the clear specifics of underground construction projects, FIDIC came to an agreement with the ITA on the preparation of a new sample form of contract to cover tunnelling projects (or underground works in general). Currently, sources are being prepared for assignment and a workgroup is being put together to prepare the model. The FIDIC approach to unforeseeability is described below.

In the CONS/1999 Red Book at Sub-Clause 1.1.6.8, the term ‘unforeseeable’ is defined as ‘not reasonably foreseeable by an experienced Contractor by the date for submission of the Tender’. As such, the definition is left general while it is assumed that it will be necessary to evaluate every particular situation on a case-by-case basis. FIDIC fails to define the term ‘unforeseen’. This term evokes a kind of subjective negligence, omission and/or phenomenon that has already been encountered. This raises the question: where are the limits of what can be foreseen subject to the reasonable costs incurred and within the span of time for preparing the bid by an experienced contractor? Situations do of course appear that are absolutely unforeseeable or that could have been foreseen by chance only, by an exceptional expert, and so on.

One of the key differences between the P&DB/1999 Yellow Book and the EPC/1999 Silver Book is in the risk allocation of unforeseeable physical conditions.

P&DB defines the term ‘unforeseeable’ in the same way as CONS. EPC does not define the term at all, assuming the definition to be redundant, as unforeseeable risks are commonly allocated to the contractor.

Given a particular situation, unforeseeability must be evaluated in respect of time for completion and the statistical frequency of event occurrences according to historical records. The following example is quoted in The FIDIC Contracts Guide, (2000) at Section 274. If the contract completion period is three years, then an experienced contractor should have foreseen an event that tends to appear once in six years, but an event appearing once every ten years is deemed unforeseeable. For details, see the charts in Appendix E.

Another vital aspect of the foreseeability issue is site data provided to the contractor as required by the Sub-Clause 4.10 of CONS and P&DB. Under this Sub-Clause, it applies that the employer shall, before the Base Date (28 days prior to the latest date for submission of the tender), provide the contractor with information and all relevant data available to them dealing with the sub-surface, on-site hydrological and geological conditions and environmental aspects. On this basis, the employer shall also make available to the contractor all such data that they acquire after the base date. Responsibility for interpretation of all this data rests with the contractor. To the extent practicable (taking into account cost and time), the contractor shall be deemed to have obtained all necessary information as to risks, contingencies and other circumstances which may influence or affect the Tender or Works. To the same extent, the contractor shall be deemed (prior to submitting the tender) to have inspected and examined the site, its surroundings, the above data, other available information and to have satisfied themselves with all relevant matters, including (without limitation):

• the form and nature of the site, including sub-surface conditions;
  
• the hydrological and climatic conditions;
  
• the extent and nature of the work and goods necessary for the execution and completion of the works and the remedying of any defects;
  
• the laws, procedures and labour practices of the country;
  
• the contractor’s requirements for access, accommodation, facilities, personnel, power, transport, water and other services.

As can be seen, risk allocation is again shared. Risk is allocated to the contractor only to the extent that is reasonable and proportionate to the cost of the bid (e.g. site investigation, surveys) and time available for preparation of the bid.

Conversely, Sub-Clause 4.10 of EPC/1999 Silver Book transfers risk fully to the contractor. EPC stipulates that the employer must, before the base date, provide the contractor with all information and significant data available to them (relating to on-site hydrological and geological conditions) and environmental aspects. Likewise, the employer shall also make available to the contractor all such data acquired after the base date. Responsibility for verification and interpretation of all data rests with the contractor. In other words, the wording in the Silver Book contains the contractor’s obligation to verify the data. Furthermore, the employer bears no responsibility for the accuracy, adequacy or completeness of such data except for the data and information provided by the employer as per Sub-Clause 5.1 of the Silver Book:

• portions, data and information which are stated in the contract as being fixed or the responsibility of the employer;
  
• definitions of intended purposes of the Works or any parts thereof;
  
• criteria for the testing and performance of the completed Works; and
  
• portions, data and information which cannot be verified by the contractor, except as otherwise stated in the contract.

It is common practice for employers, for various reasons, to carry out site inspections and investigations (including geological surveys) at the tender stage. This is to accommodate for budget contingencies, to prepare the tender design, to assess the criteria for risk evaluation, and so on. There may also be a requirement upon the lender to conduct a respective risk analysis, feasibility study, and so on. At the tender stage, bidders need all such information gained by the employer to enable them to foresee (as much as possible) the work conditions and physical conditions that might affect implementation.