Producing reliable and robust prognoses on cutter wear, drilling progress and related costs are becoming an increasingly sensitive issue for machine manufacturers and contractors dealing with operation of mechanical excavation systems in tunnelling, underground construction and mining. Equally important for the clients and owners is that cost and time estimates must have adequate quality. Representative and trustworthy parameters describing the strength and abrasive properties of rock and soil are crucial as these constitute the fundamental input for obtaining reliable cost estimates. This is equally important when it comes to risk assessments described by terms such as ‘low’, ‘good’, etc. in tender documents, and in situations where claims are filed during or following the construction period.

A main strength of the NTNU/SINTEF drillability tests is that the test results may be put into prediction models to estimate time and costs of rock and soil excavation projects. Other commonly used rock test methods do not have prediction models of the same extent available.

Recently several laboratories have set up equipment to perform the NTNU/SINTEF drillability tests. This has in some cases given highly unexpected test results; unexpected that is partly as substantially diverging results of parallel testing between other laboratories and the NTNU/SINTEF laboratory, and partly as results that are systematically outside the variation of the test results recorded in the NTNU/SINTEF database for the actual rock types.

Hence NTNU/SINTEF is concerned that incorrect test equipment, procedures or results may discredit the test methods and related prediction models.

Background
There are several laboratory test methods that can be used to determine properties influencing the drillability of rock. It is however essential to use consistent, comparable and reliable test procedures in order to achieve solid and robust prognoses and cost estimates.

The Brittleness Value (S20), Sievers’ J- miniature drill test (SJ), Abrasion Value (AV) and Abrasion Value Cutter Steel (AVS) have been extensively used at NTNU/SINTEF since the 1960s in connection with drillability testing of rock samples. The Drilling Rate Index (DRI), Bit Wear Index (BWI) and Cutter Life Index (CLI), commonly referred to as the NTNU/SINTEF drillabilty indices, are all calculated on the basis of the above mentioned individual test methods.

Samples originating from more than 3000 projects in more than 50 countries have so far been tested at the NTNU/SINTEF laboratory. Performance prediction and cost evaluation models for drill-and-blast tunnelling, TBM tunnelling and rock quarrying have been developed by correlating laboratory and in situ geological data with production data from tunnel projects. The models are continuously updated and revised as new tunnelling data become available. The NTNU/SINTEF method is today a recognised and widely used test method for, for example, TBM performance prediction.

The increased demand for knowledge on properties influencing the time and cost of tunnelling has instigated further developments of the test methods. Until the development of the NTNU/SINTEF SAT [Nilsen et al, 2007,1], very limited knowledge was available to describe the abrasiveness of soil and its impact on soft ground TBMs. The SAT is a further development of the AVS test for rock samples. The SAT method has, since the introduction in 2006 [Nilsen et al, 2006,2], already been used for more than 20 different tunnelling projects, and around 200 samples have so far been tested at the NTNU/SINTEF laboratory for determination of SAT.

Implications
The laboratory test equipment used by NTNU/SINTEF is not commercially available. The principles and details of the tests are widely published, and duplicates of equipment can be established based on this information. As mentioned, several laboratories have decided to establish such replicas for performing the suite of tests required to determine the NTNU/SINTEF drillability indices.

Experience shows that these replicas may produce results that are not comparable and consistent with the results from the NTNU/SINTEF reference laboratory. The main source of inconsistency is most likely related to various differences in equipment design, manufacture and installation, and in the materials used. The inconsistency in test results is also related to the fact that, although the principles of the tests may seem straightforward, the preparation, and testing itself, involve several critical steps and operations which can significantly affect the obtained the results [Dahl, 2003,6].

Hence NTNU/SINTEF cannot approve the quality of tests done by other laboratories without performing extensive calibration and parallel testing. Currently NTNU/SINTEF has not established a system for such calibration testing.

The problem with obtaining comparable, reproducible and consistent test results coming from different laboratories is not unique to NTNU/SINTEF test methods. For the Cerchar Abrasiveness Index (CAI), one of the most commonly used tests to evaluate rock abrasiveness for cutter life/cost estimates of TBM projects, several studies have shown that technical factors can have a significant impact on the results [Plinninger et al, 2003,3], and that the many different approaches used at various laboratories lead to results which are not consistent and directly comparable [Rostami et al, 2005,4]. Hence, it is important for the end users of the CAI to be aware of these variations and their impact on cutter and bit cost estimates [Rostami, 2006,5] .

NTNU/SINTEF frequently receive inquiries regarding the apparatus and the test method itself. Detailed information and suggested standard procedures are therefore made available at the NTNU/SINTEF website: www.drillability.com. It should however be emphasised that NTNU/SINTEF do not intend to try to approve test equipment, test procedures or test results of other laboratories. Therefore, NTNU/SINTEF can not, even when a comprehensive parallel testing program has been executed, take any responsibility or liability for the reliability and quality of test results and calculated drillability indices originating from laboratories other than the NTNU/SINTEF reference laboratory.

Conclusions
NTNU/SINTEF is committed to ensure end users they will get quality assured and reliable drillability test results, which can be used to obtain the best possible time and cost estimates. Consistent and repetitive testing is important to ensure that all end users have confidence in these quantitative methods for describing drillability characteristics of rock and soil materials. This is the main reason for labelling the NTNU/SINTEF indices as trademarks. It is also an intention to try to avoid some of the ambiguities that may be associated with other methods, e.g. the CAI.

The NTNU/SINTEF prognosis models are along with the extensive and constantly increasing test database unique, and constitute a useful asset for cost and time estimates, risk and uncertainty analyses, classification of test results and correlation with other test methods when utilising ™ labelled test results DRI™, BWI™, CLI™ and SAT™.


Outline of the brittleness test used to determine Brittle Value (S20) Brittleness testing for determination of S20′ close-up of impact weight hitting mortar with rock sample material [Photo: Simon A Hagen] Outline of the sievers’ J miniature drill test to determine Sievers’ J-Value (SJ) Miniature drill testing for determination of SJ Value, Close-up of miniature drill bit and rock specimen Outline of the abrasion test used to determine Abrasion Value (AV), Abrasion Value Cutter Steel (AVS) and Soil Abrasion Test (SAT) Abrasion testing for determination of AV/AVS/SAT. Close-up of test piece running on prepared abrasion powder [Photo: Simon A Hagen]