The North American Tunneling Conference held in Atlanta this April, featured an all-day workshop on Geotechnical Baseline Reports. Chaired by Bart Bartholomew, the session involved commentary by a panel of engineers, owners, contractors and legal representatives from the US tunnelling industry.

The focus of the workshop was to address improvements that could and should be made in the preparation, interpretation, and application of Geotechnical Baseline Reports (GBR). Each panellist was given the opportunity to address the three most pressing concerns they had with regard to GBRs, with a designated “alter ego” on the panel responding to each issue. Audience participation was also encouraged.

In all, 18 separate issues were raised and responded to by the panellists during the day-long session, and at least a dozen more were raised through audience commentary. In several instances, distinctly divergent opinions existed, particularly on matters of legal interpretation and the manner in which baselines are explained and expressed. In many cases, however, panel members concurred in their views.

The evolution and use of GBRs

The use of a geotechnical interpretive report was first incorporated into a US construction contract in 1972, on the Washington DC subway. In the subsequent 32 years, the use of these reports has slowly but progressively evolved into what are now termed Geotechnical Baseline Reports.

These documents have been the subject of a number of publications, the most recent being a 1997 publication prepared by the Underground Technology Research Council (UTRC) within the American Society of Civil Engineers, entitled “Geotechnical Baseline Reports for Underground Construction, Guidelines and Practices.” One of the workshop panellists, and co-author of this article, Randall Essex, chaired the UTRC subcommittee that coordinated industry feedback sessions, and was principal author of the Guidelines.

Over the years, the use of a GBR as a risk identification/risk sharing tool has gained increasing acceptance. Most recently, the GBR has been cited in the “Joint Code Practice for Management of Tunnel Works in the UK” (September, 2003), by the Association of British Insurers and the British Tunnelling Society, as a standard for the representation of anticipated subsurface conditions in a contract prior to bidding.

Where can we improve?

While the panellists’ and audience opinions indicated they support the concept of a GBR as an important risk management tool, various visions emerged regarding its content, its form of presentation, authorship, internal coordination with other contract documents, and interpretation within legal doctrines.

As the day progressed, it became clearer what is working, what is not, and what might be improved/ modified to enhance the effectiveness of GBRs. A broad array of issues were addressed, many focusing on real “nitty-gritty” topics that have arisen through the resolution of disputes on past projects. A sampling of the issues discussed follows:

Are baselines only enforceable if based solely on the available data?

It was discussed that in many instances the GBR database may be neither complete nor entirely representative of the anticipated subsurface conditions. Examples cited included boulders not detected in a soil boring programme and anticipation of vertical shears or faults missed due to absence of inclined borings.

Added focus in the text of the report is needed to explain when and why the baselines deviate from the available database. In some cases, the explanation could result from experience on a nearby project that is beyond the actual site of the project being tendered. It was generally concluded that properly stated and explained baselines should be considered binding as contractual statements regardless of the presence or absence of substantiating databases.

Should DRBs respect the baselines?

It was offered that a body of lessons-learned decisions resulting from Dispute Review Boards (DRB) over the years has created an informal “case law” that can be influential in guiding future practice. It was expressed that a DRB certainly carries a formal responsibility to the parties involved to enforce the baselines and terms of contract, but may also carry an informal responsibility to the industry; to provide recommendations that will sensibly guide future practice.

It was suggested that in a competitive bidding environment, baselines protect three key entities – the owner, the low bidder, and the second low bidder. If dispute resolutions (through DRBs or other means) do not respect the baselines, owners will lose faith in the process and contractors will tend to ignore the risks allocated in the baseline statements and provisions.

Perhaps fairer, more robust pricing of the work, with fewer attendant claims, is the best way forward for the industry. In order for this to evolve, DRBs must be diligent in administering the baselines as stated.

Should baselines be precise values, or can/should they be ranges?

The panel was in substantial agreement that “soft” baselines, which express a broad range of material properties or characteristics, run counter to the GBR philosophy – to provide clarity and eliminate ambiguity. A baseline “range” could be interpreted in a manner that serves neither objective. In fact, before the development of GBRs, only raw geotechnical data was provided – in the form of ranges. Thus the stating of baselines as a range, appears to be a step backward, not forward.

Several in the audience expressed a concern about being “wrong” if an average property or characteristic was presented as a baseline. This important concern resulted in two different responses.

First, while a nominal difference in material property or characteristic might provide suitable evidence to support a claim, the actual costs owing to such a nominal difference would also be nominal. By paying such nominal claims the baseline concept is honoured, and focus returns to the work at hand. On the other hand, if an owner (or owner’s agent) was concerned that presenting an average condition as a baseline would lead to a number of (minor) claims, they could reasonably adjust the baseline upward – recognising that in exchange for a more conservative baseline, they could anticipate higher bid prices. Adjusting the baseline to a more conservative position was viewed as a better response to an owner’s fear of being “wrong” than presenting a baseline as a range.

Can a bidder interpret the contract payment quantities as a baseline?

The panel was generally in agreement that planned quantities in a bid schedule were fair indications of anticipated conditions, and that baseline statements made in the GBR should be compatible with such payment provisions. This is precisely the clarity and consistency that is sought in the proper preparation of a GBR – compatibility with payment provision statements and quantities will serve to eliminate ambiguity, confusion, and disagreement.

Are unreasonably onerous/conservative baselines binding, or can they be ignored?

The panel recognised that some DRB decisions have supported the conclusion that if baselines are unreasonable and lacking a basis for departure from the available information, they may be set aside in favour of the data. Whether this honours the concept that “the baseline is the baseline” was also debated.

The conclusion was those preparing GBRs should respect the importance of being as clear as possible regarding the basis of the baselines presented. Stating unrealistically conservative or adverse baselines tends more toward a position of “risk shedding” than “risk sharing”, which defeats the general use of a GBR as a risk-sharing tool. It was generally recognised that an entire day’s debate could have been dedicated to this topic, and that there was no 100% correct answer.

How large a variance is required from the baseline to justify additional compensation?

This question also drew extended debate. It was acknowledged “it depends” on how the baselines are stated. For instance, if a baseline indicated that five fault zones were to be expected and six were encountered, this would represent a change or variation subject to additional compensation.

However, if a baseline indicated the average UCS of rock to be bored with a TBM could be expected to be 172MPa, and the measured average during the drive was 175MPa, would this be a fair and reasonable basis for a claim? The concept of what constitutes a “material difference” in the contract definition of a Differing Site Condition (DSC), whether physical or behavioural, would need to enter into the discussion.

Does a contractor have to demonstrate reliance on a baseline in order to recover damages?

The attorneys on the panel disagreed on this point. Before GBRs were used, the legal doctrine in US contracting practice was that a bidder needed to show that they “reasonably relied” on a contract indication that was found to differ in order to secure relief under the DSC clause. If it was not important at the time of bid, why was it important now?

However, one panellist reasoned that because GBRs stipulate baselines, a DSC claim can now be determined solely on the basis of whether conditions differ from the baselines, regardless of reliance. No resolution was reached on this matter, which might suggest that additional commentary from the legal community is warranted.

Can the GBR process be adapted for Design-Build?

It was acknowledged that the UTRC Guidelines were developed within the context of traditional design-bid-build contracting only. The panel discussed a modified approach, where a more collaborative exchange would occur between the owner (or owner’s agent) and the design-build team. Attention was directed to a chapter in a recent publication by the American Underground Construction Association that addresses geotechnical reports for design-build project delivery (Design-Build – Subsurface Projects, Brierley & Hatem, AUA, 2002).

Should the Geotechnical Data Report also be included in the construction contract?

This topic also drew spirited debate. One perspective was that since the GBR was the primary geotechnical document for assessing the existence of a DSC, why muddle the contract with an additional Geotechnical Data Report. The other perspective offered was that since the GBR could not be expected to be exhaustive in the description of “every possible baseline condition under the sun”, that the available data should be accessible to the parties as a contract document in order to address instances when the GBR might be silent regarding a particular circumstance.

What if the GBR is silent on a particular condition?

One opinion was that there needed to be a measure of reasonableness applied to both the preparation of a GBR and it’s interpretation. If a ground type is stated to be unstable, and this instability could be manifested in a number of described “misbehaviours”, does the fact that GBR overlooks one of those misbehaviours constitute a DSC? Should the GBR be held to a standard of having to express every potential behaviour “under the sun”?

The panel and audience generally concurred that this was not a fair standard to apply, particularly when addressing ground behaviour, which can be affected by aspects such as equipment, workmanship, procedures, and worker skill. However, if a GBR fails to indicate, for instance, the occurrence of heavy faulted ground in an otherwise competent rock tunnel drive, and that condition was encountered, would a claim be justified? The general consensus was “yes”.

The way forward

All in all, the workshop was very successful. While the panellists and attendees appeared to support the purpose and role of GBRs in the contracting process, there was healthy debate regarding the details of implementation. One strong conclusion was that when the authors of such documents do a better job anticipating conditions that are likely to occur, and write clear and concise baselines to describe those conditions, the effectiveness of GBRs is enhanced.

Based on a show of hands at the end of the workshop, the attendees’ opinion was that an update of the UTRC GBR Guidelines would be beneficial to clarify the types of issues that were raised during the session.

The authors will carry forward the recommendation to the UTRC that a second edition of the GBR Guidelines be developed and made available to the industry. Readers should look forward to more conference sessions of this nature, where opinions can be heard and reflected in improved guidance documents.