MAEDA KENSETSU KOGYO KABUSHIKI KAISHA also known as MAEDA CORPORATION AND ANOTHER v. BAUER HONG KONG LTD

CACV 113 & 122/2019

(Heard together)

[2020] HKCA 158

IN THE HIGH COURT OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

COURT OF APPEAL

CIVIL APPEAL NOS 113 OF 2019 AND 122 OF 2019

(ON APPEAL FROM HCCT NO 35 OF 2016)

________________________

______________________

BETWEEN

______________________

(Heard together)

______________________

J U D G M E N T

______________________

Hon Kwan VP:

Introduction

1.  There are two appeals before this court from the judgment of Anthony Chan J on 18 May 2018 (“the Judgment”).  For ease of reference, the expressions and terms used in the Judgment will be adopted.

2.  The judge heard an appeal on several questions of law arising out of the First Interim Award (“Award (1)”) of Sir Vivian Ramsey QC (“the Arbitrator”) published on 21 June 2016 in respect of an arbitration between the plaintiffs (“the JV”) and the defendant (“Bauer”). The questions concerned the proper interpretation of contractual Measurement Rules (“MRs”) M12A to M12G, which were made for the measurement and payment of the excavation of diaphragm wall (“D-wall”) panels.  In granting leave to both parties to appeal from the Judgment to the Court of Appeal, the judge noted that the construction exercise he carried out was particularly difficult due to the poor drafting of the MRs^[1].

3.  The JV issued the originating summons in these proceedings (HCCT 35/2016) on 21 July 2016, seeking leave to appeal to the Court of First Instance with four grounds of appeal^[2].  Leave to appeal was granted by Deputy High Court Judge Lee (as Lee J then was) on 7 April 2017, pursuant to section 6(4) of Schedule 2 to the Arbitration Ordinance, Cap 609^[3]. The appeal was heard by Anthony Chan J over a three-day hearing.  As required by section 5(4) of Schedule 2 to Cap 609, the judge determined the questions of law ab initio.

4.  After the granting of leave to appeal by DHCJ Lee, the Arbitrator published the Second Interim Award on 3 January 2018, in which he applied his earlier interpretation of the MRs.  The JV issued an originating summons in HCCT 6/2018 on 1 February 2018 seeking leave to the parties to appeal against that award as well and an order that the later proceedings be heard together with HCCT 35/2016.  There was agreement that further awards may need to be revised depending on the outcome of HCCT 35/2016.  After the Judgment was given and on 27 February 2019, the judge ordered that any consequential, additional or corrected award or order of the Arbitrator be remitted to the Arbitrator for reconsideration in light of the Judgment.

Background

5.  I adopt gratefully the background matters set out comprehensively in §§2, 8 to 19 of the Judgment and annex to this judgment the same annexures as in the Judgment: Annex I (being items C2(a) and C2(b) of the Preamble of the Pricing Document, the MRs are contained in item C2(b)); Annex II (two sketches to illustrate some of the explanations, taken from attachment C to the skeleton submissions of Mr Peter Clayton SC, who appeared for the JV here and below); and Annex III (pages 100 to 103 of Award (1), which contained the Arbitrator’s decision on the construction of the MRs):

“2. The JV were the main contractors employed by the Mass Transit Railway Corporation (“MTRC”) under 2 main contracts to construct tunnels for the Hong Kong to Guangzhou Express Rail Link, namely, Contract 823A and Contract 823B.  These Contracts were administered on behalf of MTRC by an Engineer whose role and duties were set out in the Contracts.  The JV sub-contracted the D-wall works under both Contracts to Bauer under Sub-contracts 823A and 823B.  Award (1) relates to the parties’ disputes under Sub-contract 823B.”

“8. D-walls are concrete walls constructed in trenches excavated into the ground using a series of interconnected panels. Before excavation, drill holes were undertaken along the line of the D-walls to investigate the ground and recover core samples.

9. The Measurement Rules (“MRs”) were made specially by MTRC for the main contract. They were passed onto Sub-contract 823B for the measurement and payment of the excavation works. To put them in proper context, the MRs formed part of the Preamble (a 73 page document) of the Pricing Document which was prepared based on the Civil Engineering Standard Method of Measurement (3^rd edn) (also known as CESMM3) published for the Institute of Civil Engineers. In particular, the Preamble set out the departures from CESMM3 in accordance with para 5.4 thereof.

10. It should be noted that para 5.5 of CESMM3 provided that: “Where excavation, … is included in the work a definition of rock shall be given in the Preamble and this definition shall be used for the purposes of measurement.”

11. The MRs came under Part A (Amendments to CESMM3), Class C^[4] (Geotechnical and Other Specialist Processes) of the Preamble. Under item reference C2 (Diaphragm Walls), there was an amendment to Second Division and Third Division of the work classification under Diaphragm Walls. To understand the work classification, one needs to refer to para 3.1 of CESMM3:

“The Work Classification divides work commonly encountered in civil engineering contracts into 26 main classes.  Each class comprises up to three divisions which classify work at successive levels of detail.  Each division comprises a list of up to eight descriptive features of work.  Each item description in the Bill of Quantities shall identify the component of work covered with respect to one feature from each division of the relevant class, …”

12. Under item C2(a), Diaphragm walls (First Division) were divided into “Excavation in material Other than rock m³” and “Excavation in rock m³” under Second Division. The latter was further divided under the Third Division into 3 categories: “Rock – Class A”, “Rock – Class B” and “Toe in to rock”.

13. Bauer was entitled to be paid for the work involving excavation in 4 types of identified material at different rates per m³ as priced in the Sub-contract Schedule of Prices: “material other than rock” (“Other Material”) at HK$X; both “Rock – Class A” and “Rock – Class B” at HK$Y; and “Toe in to rock” (“TITR”) at HK$Z. Self-evidently, the pricing reflected the difficulties in the works.

14. It is common ground that the MRs were used by the parties for the tendering under Sub-contract 823B.

15. The MRs were set out in 5 pages under item C2(b), a copy of which together with C2(a) is attached hereto as Annex I. It is not at all easy to understand them, and in order to do so (apart from the foregoing introduction) one needs to have an understanding of the relevant features of the works and the provisions under the Sub-contract. For this purpose, I am grateful to Mr Clayton for having set them out in succinct terms as part of Attachment A of his skeleton submissions. The relevant paragraphs are 1 to 11, with the exception of para 8(2)^[5] which consists, at least in part, of submissions. In addition, there are 2 sketches provided in Attachment C to illustrate some of the explanations. I believe that these materials are accurate^[6] and helpful to gaining a better understanding of the relevant background for the interpretation exercise in hand. Attachment C is attached hereto as Annex II and the relevant parts of Attachment A are set out below (footnotes omitted). For ease of reference, the paragraphs are renumbered from 15.1 to merge with this judgment:

The nature of the works

15.1 D-walls are reinforced concrete walls constructed deep into the ground using a series of interlinked panels and specialist machinery. In the present case parallel D-walls, on either side of the proposed railway tunnel alignment, were constructed by [Bauer] under Sub-contract 823B in two areas of the project known as the Emergency Rescue Sidings (‘ERS’) and Cut and Cover tunnel (‘C&C’) sections. In summary, the process of D-wall panel construction is as follows:

(1) excavation of a deep trench for each panel, typically 6.0m or 6.3m long by 1.0m or 1.2m wide, along the alignment of the D-wall. The excavation is carried out using grabs and chisels and/or cutting machines.

(2) the trench excavation stops at a founding level determined by Engineer in accordance with relevant founding criteria and verified during excavation. This may or may not require going into sound rock depending on the relevant founding criterion.

(3) a reinforcement cage is lowered into the excavated trench.

(4) concrete is poured into the trench. Upon concreting the D-wall panel is, in principle, complete.

(5) the process continues along the alignment of the D-wall.

15.2 The [JV] were to excavate in between the completed D-walls to below the intended track level of the rail lines. The D-walls were to retain the ground on either side of and prevent the ingress of water into the excavation and in the ERS section formed part of the final structure built by the [JV]. In the C&C section the D-walls were temporary works allowing construction of a tunnel section in between them prior to backfilling of the excavation.

15.3 In both the ERS and C&C sections, base slabs were to be constructed by the [JV] at the bottom of the excavations between the D-walls on which the rail tracks would run, as can be seen from figure 1 in [Annex II] which is taken from one of the Sub-contract drawings.

The founding criteria and other relevant provisions

15.4 A very important consideration in the design and excavation of the D-walls is the engineering founding criteria for the D-wall panels. In other words, the criteria for determining the material into which the D-walls should be toed (keyed) or founded on and their depth.

15.5 (1) The D-wall engineering founding criteria were given on the Sub-contract drawings. For instance, Drawing No.823B/T/000/OAP/C01/004B:

1. All permanent diaphragm walls shall satisfy either of the following founding criteria:

i) Should the category 1(c) rock head be higher than the base slab bottom level, the diaphragm wall shall be founded on at least 300mm below the base slab bottom level and shear pin shall be provided according to the diaphragm wall schedule.

ii) Should the category 1(c) rock head be less than 15m below the base slab bottom level, the diaphragm wall shall be founded at least 300mm into category 1(c) rock head and shear pin shall be provided according to the diaphragm wall schedule.

iii) Should the category 1(c) rock head be more than 15m below the base slab bottom level, the diaphragm wall shall be provided with a minimum length of 15m below base slab bottom level and be founded on category 1(d) or better rock (emphasis added)

The final formation level is assumed to be not more than 100mm below base slab bottom level.

The founding material shall satisfy the requirements as stated in the following schedule.

Table 1 founding material schedule

Rock material decomposition grade	Minimum uniaxial compressive strength UCS (MPa)	Equivalent point load index strength PLI50 (MPa)	Total core recovery	Allowable bearing pressure (kPa)
Category 1(c) or better	25	1 MPa min.	85% of the specified grade min. calculated from a 1.5m core	5000
			run
Category 1(d)	-	-	50% of the specified grade min. calculated from a 1.5m core run	3000

(2) It can be seen from the three engineering founding criteria above that the critical element for determining founding levels is the level of “category 1(c) rock head” and that category 1(d) rock for founding criterion (iii) is not expressed as being a “rock head” in the founding criteria.  Thus, according to founding criteria (i) and (ii), the D-walls are required to be founded into the “category 1(c) rock head”, whereas under founding criterion (iii), the D-walls are only required to be founded on category 1(d) rock.

(3) Table 1 of this Drawing stipulates total core recovery from a 1.5 metre core run for the material to be accepted as category 1(c) is required to be 85% minimum of the specified grade and also with a minimum uniaxial compressive strength UCS of 25.  For category 1(d) the requirement is only 50% of the specified grade and without any minimum uniaxial compressive strength.

(4)(i) A description of category 1(c) and 1(d) rock is to be found as Table 2.1 of the Code of Practice for Foundations, which is reproduced, so far as category 1(d) is concerned, at paragraph 500 of [Award (1)].  The relevant descriptions of the categories are as follows:

1(c) Slightly to moderately decomposed moderately strong rock of material weathering grade III or better, with a total core recovery of more than 85% of the grade and minimum uniaxial compressive strength of rock material (UCS) not less than 25 MPa (equivalent point load index strength PLI₅₀ not less than 1 MPa)

1(d) Moderately decomposed, moderately strong to moderately weak rock of material weathering grade better than IV, with a total core recovery of more than 50% of the grade

(ii) Classification of rock weathering or decomposition grades I to VI (as referenced in the description of categories 1(c) and 1(d) above) can be seen from Table 4 in Geoguide 3.  Grade I is fresh rock with grades II to VI reflecting progressive stages of decomposition down to residual soil.  The 85% of the specified grade (category 1(c)) and 50% of the specified grade (category 1(d)) as mentioned in Table 1 (paragraph [15.5 (1)] above) is accordingly of grades I to III.

(5) A simple diagrammatic illustration of the different engineering founding criteria for Sub-contract 823B appears in figure 2 in [Annex II].

15.6 Core recovery is a reference to core samples obtained from the drillholes carried out prior to excavation.  Thus, the key distinguishing feature between category 1(c) and 1(d) is that 1(c) will have 85% of the length of core comprising rock of grades I to III, whereas 1(d) need only have between 50% and 85% of the length of core comprising rock of grades I to III.  Additionally, category 1(d) need not satisfy the minimum uniaxial compressive strength requirement.

15.7 One of the main reasons why there are different founding criteria is that the natural ground conditions are variable. Whilst generally less weathered rock is found as the excavation proceeds down, there are frequently pockets of more decomposed material within the weathered rock and rising and falling profiles of category 1(c) and category 1(d) material.  As part of the Contract 823B works, the [JV] drilled small diameter drillholes along the lines of the D-walls which were used to obtain core samples and undertake tests in relation to the material to be excavated.  From the results of the drillholes the relevant category 1(c) rock head and 1(d) rock levels at the drillhole location could be established and, by applying the relevant engineering founding criteria, tentative founding or toe levels for individual panels were determined and provided to [Bauer] for its works.  For the panels as excavated, the founding criteria (as set out above) require the category 1(c) or 1(d) material to be present across the whole panel base at that level and below, as shown in [Annex II].

15.8 (1) Whilst the carrying out of drillholes reduced some of the uncertainty as to the rock profiles prior to excavation of the panels, the coring equipment used for the drillholes was generally only between around 60-85mm in diameter.  Most of the D-wall panels were between 6m and 6.3m long by 1m or 1.2m wide in the horizontal plane.  For a 6m by 1m panel that means the drillhole is covering less than 0.1% of the area of a panel at any given level.  There would likely still be considerable variability in the rock profile within panels and between individual drillholes and it is very unlikely that an individual drillhole will hit the point where category 1(c) material is present across the whole base of a particular panel, rather than a higher point where category 1(c) is present across some of the panel.

15.9 The final excavated founding or toe levels for the panels were verified for approval by the Engineer on site during excavation, so the rock head level across the whole panel and, hence, the founding or toe level could be checked and identified.

15.10 Measurement Rule M12D provides [inter alia] that:

“Bedrock” shall be rock below the founding level as determined by the Engineer in accordance with the M&W Specification

15.11 The M&W (Material and Workmanship) Specification, part of Contract 823B, provides a definition of rock at clause 24.2(15) and bedrock at clause 24.2(16) thereof:

Rock is naturally occurring material of Grades I to III as classified in Table 4 of Geoguide 3 “Guide to Rock and Soil Descriptions”

“Bedrock” means rock which is defined by description to have been an originally solid in place rock mass below which Grade IV, V or VI material normally would not or has not been found.

16. The above extraction of Attachment A referred to “rock head” and “Geoguide 3”. There is a definition for the former in Consultancy Agreement No C803, which was one of the contractual documents of Contract 823B.  It set out the geotechnical conditions anticipated to be encountered during underground construction.  Paragraph 7.1 of that document contained a glossary of geological terms where “Rockhead” can be found, and it was defined as “the surface of the Bedrock”.  Although there may be difference between the parties on the relevance of this document, there is no disagreement that this is a correct definition.

17. Geoguide 3 is a document produced by the Geotechnical Engineering Office of the Hong Kong Government.

18. In addition to the above background, it is important for the understanding of the MRs to bear in mind that, firstly, underground condition varies, eg, even when one has excavated to a layer of rock, nature is such that it is highly unlikely for the depth of that layer to be uniform.  Secondly, underground condition is not fully known until the ground is opened up.  Thirdly, the MRs served, in part, the purpose of allowing the contractors to tender for the works without full knowledge of the underground condition. 

19. Fourthly, the measurement exercise was never meant to be a precise one.  This is self-evident from the MRs.  One of the key constituents of the MRs is the “measurement horizon”, which is deemed to be at a fixed height of 1.5m above the highest point of the “theoretical bedrock line”.  Another example is the fact that TITR is measured at a fixed rate irrespective of the actual amount of rock or the combination of material excavated in that zone.  In the context of a construction contract, such rough and ready measurement is not uncommon and the logic can be easily understood.  It is inherent in such a methodology, or perhaps in the nature of excavation work, that there is a risk that the underground condition may turn out to be less favourable than assessed (and vice versa).”

The scheme of the MRs

6.  As noted by the judge, the MRs were poorly drafted.  The use of technical terms adds to the difficulty in understanding them properly, the similarity of some of the technical terms gives rise to unnecessary confusion, and the lack of clarity in one concept is multiplied when inter-related concepts contained in a Rule are construed^[7].

7.  The MRs are numbered from M12A to M12G.  We are concerned with M12A to M12F.

8.  To recap, the general scheme of the MRs is that the four types of excavation are to be paid for at different rates, save that Rock Class A and Rock Class B would be paid at the same rate.  For the purpose of payment, some rough and ready assessment rules were devised^[8].  The central feature of the scheme is the “measurement horizon” (“MH”), an artificial concept devised under M12D and M12E for the rough and ready assessment.  In cases where the panel was “toed into bedrock”, the excavation above the MH would be paid for on the basis of Other Material, Rock Class A and Rock Class B, whereas the excavation below the MH would be paid for as TITR.  In cases where the panel was not toed into bedrock, the excavation would be paid for only as Other Material, Rock Class A and Rock Class B^[9].

9.  M12A is a diagram with legend called Diagram C1, and is provided for illustration purpose.  M12B contains the definitions for the four types of excavated material. The definitions of Rock Class A and Rock Class B refer to “measurable core” (“MC”) as described in M12C. As stipulated in M12B, the length of the MC for Rock Class A is more than 200 mm but not exceeding 1,300 mm, and the length of the MC for Rock Class B is exceeding 1,300 mm.  M12C specified two requirements in the definition of MC: a total core recovery of at least 85% specified rock material (ie category 1(c) rock) and a rock quality designation (“RQD”)^[10] of at least 30.  Category 1(c) rock might or might not satisfy the requirement of RQD ≥ 30, so the latter is additional. 

10.  TITR is defined in M12B as “that rock between the [MH] and the toe level of the [D-wall] as determined or instructed by the Engineer”. The words “that rock” gave rise to much debate before us.  Another problem is to find out where the MH should be, with reference to M12D and M12E, an exercise which the judge has described as “challenging”^[11].  There is no dispute that the material between the MH and the toe level would be paid at TITR rate.

11.  M12D applies where D-walls are “to be keyed into bedrock”^[12]. It seeks to set out the parameters for working out a “theoretical bedrock line” (“TB Line”).  Two related concepts are introduced in M12D to work out the TB Line. The first concept is “theoretical bedrock level” (“TB Level”), which is defined as “the top level of the “measurable core^[13] providing that further “measurable core” of natural rock recovery from lower levels together with the information given by the adjacent “drillholes” provides evidence of bedrock”.  As quoted in the Judgment earlier, there is a definition of “bedrock” in M12D (there is dispute as to the meaning of this definition).  The second concept is “theoretical founding level” (“TFL”) (there is also dispute whether there is any such concept), in that TB Line is defined as “the straight lines joining the successive theoretical founding levels at each “drillholes” location”.^[14].

12.  The TB Line is used in turn in M12E for working out the MH.  Under M12E, the MH for each D-wall panel is deemed to be a “horizontal line sited 1.5 m above the highest point on the [TB Line] within that panel”.

13.  In respect of D-wall panels which are “not to be keyed into bedrock”, M12F applies.  This stipulates that M12D and M12E shall not be applicable and that the actual toe level of the D-wall panel shall be deemed to be the MH for measurement purposes.  Hence, there will be no TITR measured and the TITR rate would not apply in this instance.

The Arbitrator’s decision

14.  The area affected by the Arbitrator’s interpretation of the MRs is TITR, the measurement of which runs from the bottom of the excavation (which is uncontroversial) to a level determined by reference to the top of the MC.  The question is whether the definition of the MC in M12C, with its two specific requirements (total core recovery of at least 85% specified rock material and RQD of at least 30), should apply to the measurement of TITR.  The Arbitrator took the view that it should not, this interpretation has resulted in increases of the quantity of material measured as TITR.

15.  In gist, the Arbitrator’s reasoning is that the purpose of M12C is to deal with MC where it is used to classify rock as Rock Class A or Rock Class B.  The reference to “the top level of the [MC]” in M12D is not related to the classification of the rock to be excavated as Rock Class A or Rock Class B, but is related to the concept of bedrock level as determined by the Engineer^[15]. So in the context of M12D, “measurable core” should be given its natural and ordinary meaning as a core which can be measured, being a core recovery of at least 85% for category 1(c) or more than 50% for category 1(d) as in the criteria for those categories^[16]. Hence, there is no requirement under M12D for an RQD ≥ 30.  The JV’s contention to the contrary departs from the objects of CESMM3, in particular §2.5^[17], and the measurements derived are not just artificial but are unlikely to be a commercially reasonable outcome^[18].

16.  Diagram C1 illustrates the relationship between the TB Line and the MH, and identifies the point where the TB Line intercepts each drillhole as “Founding levels determined by Engineer”.  M12D describes the way in which the TB Level at each drillhole is determined and states that TB lines are the “straight lines joining the successive theoretical founding levels” (italics supplied).  The way in which the Engineer determines the founding level in accordance with the M&W Specification depends on the application of the founding criteria, which are different to the bedrock criteria in M12D.  The Arbitrator held that in Diagram C1, as in M12D, something has gone wrong with the language because of a confusion between bedrock levels and founding levels.  He considered that in Diagram C1, the dashed lines with an arrow marked “Founding levels determined by Engineer” should correctly say “Bedrock level”, and that in M12D, “Theoretical bedrock lines” should instead be described as “straight lines joining the successive theoretical bedrock levels” (italics supplied)^[19].

17.  As for the definition of “Bedrock” in M12D (being “rock below the founding level as determined by the Engineer in accordance with the M&W Specification”), the Arbitrator rejected the JV’s contention that category 1(d) material does not fall within that definition. Category 1(d) material is defined in the Code of Practice for Foundations as “Moderately decomposed, moderately strong to moderately weak rock of material weathering grade better than IV, with a core recovery of more than 50% of the grade”.  “Bedrock” in the M&W Specification is defined as “rock which is defined by description to have been an originally solid in place rock mass below which Grade IV, V or VI material normally would not or has not been found”.  The reference to “core recovery” in the definition of category 1(d) material means that the fact that it is better than Grade IV must be established by a core recovery of more than 50% of the material and is a measure of the quality of the core.  It does not mean that a percentage of the material will be Grade IV.  He held that category 1(d) material is “Bedrock” and that “rockhead”, defined as the surface of the Bedrock, would include the surface of category 1(d) material. So when category 1(d) rockhead is less than 15 m below the base slab bottom level and at a higher level to the category 1(c) rockhead, the D-wall panel will necessarily be keyed into bedrock^[20].

18.  Hence, TITR is to be measured as part of the D-wall excavation if the D-wall is keyed into either category 1(d) or category 1(c) material^[21]. And the TB Line should be determined by reference to the drill holes and the level at which the criterion for category 1(d) or category 1(c) material was satisfied, whichever was first encountered^[22].

The grounds of appeal and the judge’s decision

19.  The three broad grounds of appeal from the Arbitrator’s decision pursued by the JV before the judge and by the parties before this court are as follows:

Ground 1

20.  Whether (a) the “Founding levels determined by Engineer” in the legend for Diagram C1 in M12A should correctly say “Bedrock level”; and (b) “Theoretical bedrock lines” in the 2^nd paragraph of M12D should be described as “straight lines joining the successive theoretical bedrock levels” instead of “straight lines joining the successive theoretical founding levels” as stated.

21.  The judge disagreed with the Arbitrator.  He took the view that the wordings of the MRs are too clear to allow any rewriting or contrary reading in this regard.  No amendment of the MRs is justified. He did not see anything wrong with using the TFL to work out the TB Line and premised upon that the MH^[23]. Bauer appealed against this holding in CACV 122/2019.  The parties were in agreement before the judge that this ground is a standalone argument.  The judge further noted that it is unclear whether the result of the appeal on this ground would affect the payment to Bauer under the Sub-contract^[24].

Ground 2

22.  Whether the definition of “measureable core” in M12C should apply to M12D.

23.  The judge agreed with the Arbitrator that the definition of MC in M12C does not apply to M12D so there is no additional requirement of an RQD of at least 30.  The judge took the view it is not altogether clear whether M12C is meant to apply to all rock and the JV’s case is difficult to accept conceptually and quite impossible to apply as a matter of practicality.  He held that there is no contractual basis to apply the RQD requirement to the work performed by Bauer and it is unworkable to do so^[25].  This is the JV’s appeal in CACV 113/2019.

Ground 3

24.  Whether “Bedrock” in the 3^rd paragraph of M12D should include category 1(d) material.

25.  The judge disagreed with the Arbitrator and held that category 1(d) material does not come within the definition of bedrock or rockhead for the purpose of M12D.  For panels which are to be keyed or toed into rock, they must be keyed into category 1(c) and not category 1(d) material. Category 1(d) material is only relevant under founding criterion 1(iii)^[26] and in those cases the D-wall needs only be founded on, not into, category 1(d) or better rock^[27].  Bauer appealed against this holding in CACV 122/2019.

26.  The judge noted that grounds 2 and 3 are intricately related because they concern the key components of M12D and affect the proper interpretation of that Rule, and that the outcome of the respective appeals by the JV and Bauer on these grounds is likely to have a significant bearing on the payment to Bauer under the Sub-contract^[28].

27.  The JV was considered by the judge as the overall winner in the appeal before him.  To reflect the fact that the JV failed in respect of ground 2, the judge ordered Bauer to pay half of the costs of the appeal to the JV^[29].

28.  Before I consider the three broad grounds of appeal in detail, I would make two general remarks.  First, it does not appear from the submissions of the parties before the Arbitrator^[30] or before this court that ground 1 is a standalone argument.  It seems to me that all three grounds are inter-connected.  The judge’s agreement with the Arbitrator on ground 2 would seem rather odd in light of his disagreement with the Arbitrator on grounds 1 and 3.  Second, it seems clear that the decision on ground 1 does have a financial effect, because if “theoretical bedrock levels” are taken rather than “theoretical founding levels” as the correct definition of TB Lines in M12D as held by the Arbitrator, this would result in a higher MH with additional material being paid for as TITR.

Relevant principles of construction

29.  I will first mention the relevant principles of construction, of which there is largely no dispute.  Both the Arbitrator^[31] and the judge^[32] have quoted extensively from relevant passages in leading judgments which laid down the legal principles on the construction of a contract.  The principles that are of particular relevance to the present case may be summarised as follows:

(1) The construction of a contract is an attempt to discover what a reasonable person would have understood the parties to have meant. This is done by identifying the meaning of the relevant words, (a) in light of the natural and ordinary meaning of those words, the overall purpose of the contract, any other provisions of the contract, the facts known or assumed by the parties at the time at the time the contract was made, and common sense, but (b) ignoring subjective evidence of any party’s intentions. (Marley v Rawlings [2015] AC 129 at §19, per Lord Neuberger of Abbotsbury, PSC)

(2) The background knowledge which would reasonably have been available to both parties in the situation in which they were at the time of the contract includes anything which would have affected the way in which the language of the document would have been understood by a reasonable man. (Investors Compensation Scheme Ltd v West Bromwich Building Society [1998] 1 WLR 896 at 912H to 913A, per Lord Hoffmann; Arnold v Britton [2015] AC 1619 at §21, per Lord Neuberger of Abbotsbury, PSC)

(3) The ‘rule’ that words should be given their natural and ordinary meaning reflects the common sense proposition one does not easily accept that people have made linguistic mistakes, particularly in formal documents.  On the other hand, if one would nevertheless conclude from the background that something must have gone wrong with the language, the law does not require judges to attribute to the parties an intention which they plainly could not have had.  If there are two possible constructions, the court is entitled to prefer the construction which is consistent with business common sense and to reject the other. (Investors Compensation Scheme Ltd v West Bromwich Building Society at 913D to E, per Lord Hoffmann; Rainy Sky SA v Kookmin Bank [2011] 1 WLR 2900 at §21, per Lord Clarke of Stone-cum-Ebony JSC)

(4) Reliance placed on commercial sense and surrounding circumstances should not be invoked to undervalue the importance of the language of the provision which is to be construed.  Except in a very unusual case, the parties have control over the language they use in a contract and must have been specifically focussing on the issue covered by the provision when agreeing the wording of that provision. Where the parties have used unambiguous language, the court must apply it.  A court should be very slow to reject the natural meaning of a provision as correct simply because it appears to be a very imprudent term for one of the parties to have agreed, even ignoring the benefit of wisdom of hindsight. Commercial common sense is not to be invoked retrospectively. (Arnold v Britton at §§17, 19 and 20, per Lord Neuberger of Abbotsbury, PSC; Rainy Sky SA v Kookmin Bank at §23, per Lord Clarke of Stone-cum-Ebony JSC)

(5) In some cases, an event subsequently occurs which was plainly not intended or contemplated by the parties, judging from the language of the contract.  In such a case, if it is clear what the parties would have intended when they entered into the contract, the court will give effect to that intention.  The role of the construct, the reasonable person, is to ascertain objectively, and with the benefit of the relevant background knowledge, the meaning of the words which the parties used. The construct is not there to re-write the parties’ agreement because it was unwise or because subsequent events have shown that the natural meaning of the words has produced a bad bargain for one side.  Commercial common sense must not be seen through the eyes of only one party. (Arnold v Britton at §22, per Lord Neuberger of Abbotsbury, PSC, and at §77, per Lord Hodge JSC; Sinoearn International Ltd v Hyundai-CCECC Joint Venture (2013) 16 HKCFAR 632 at §79, per Tang PJ)

(6) Interpretation is a unitary exercise.  Where there are rival meanings, the court can give weight to the implications of rival constructions by reaching a view as to which construction is more consistent with business common sense. But in striking a balance between the indications given by the language and the implications of the competing constructions, the court must consider the quality of drafting of the clause (the poorer the quality of the drafting, the less willing the court should be to be driven by semantic niceties to attribute to the parties an improbable and unbusinesslike intention), and it must be alive to the possibility that one side may have agreed to something which with hindsight did not serve his interest. (Wood v Capita Insurance Services Ltd [2017] AC 1173 at §11, per Lord Hodge JSC; Gan Insurance Co Ltd v Tai Ping Insurance Co Ltd (No 2) [2001] 2 All ER (Comm) 299 at §16, per Mance LJ)

(7) Where the disputed provision is open to two possible interpretations, the unitary exercise involves an iterative process, by which each suggested interpretation is checked against the provisions of the contract and its commercial consequences are investigated.  Once one has read the language in dispute and the relevant parts of the contract that provide its context, it does not matter whether the more detailed analysis commences with the factual background and the implications of rival constructions or a close examination of the relevant language in the contract, so long as the court balances the indications given by each. (Wood v Capita Insurance Services Ltd at §12, per Lord Hodge JSC; In re Sigma Finance Corpn [2010] 1 All ER 571 at §12, per Lord Mance JSC)

(8) An appellate court is entitled to take account of the fact that an experienced judge of a specialised tribunal reached the conclusion that a particular construction would have a surprising and uncommercial result.  The court dealing with an appeal against an arbitrator’s decision should approach the matter with care.  It should have full regard to the views the experienced tribunal felt able to reach on matters of business or commercial sense that form an important dimension of the context within which a decision of law is reached on the true interpretation of the contract. (Rainy Sky SA v Kookmin Bank at §41, per Lord Clarke of Stone-cum-Ebony JSC; A v B [2017] EWHC 3557 (Comm) at §13, per Knowles J)

30.  The judge did not think there is a hard and fast rule there has to be two possible constructions of a document before the court can resort to business common sense to reject one of the constructions as being unreasonable.  He was inclined to think that the court should take a holistic view of the relevant circumstances and the document to be construed, and be guided by the authorities and sound common sense^[33].  I agree with the judge.  In any event, it would appear there are two possible constructions of those parts of the MRs that are in issue.

31.  Mr Ian Pennicott SC, who appeared for Bauer here and below^[34], submitted that the judge had applied the wrong test under section 5(4) of Schedule 2 to the Arbitration Ordinance in that in respect of each of the grounds of appeal, the judge was faced with a mixed question of fact, with much technical content, and law.  The test in this instance is not simply whether the Arbitrator was incorrect as a matter of law, but whether his determination was outwith the range of solutions available to him.  In support of this, Mr Pennicott cited Arbitration Law by Robert Merkin (2004 ed) at §21.8, under the heading of “Questions of mixed fact and law”, and The Mathew [1990] 2 Lloyd’s Rep 323 at 326, per Steyn J.

32.  Some of the examples where the aforesaid test was applied are: The Mathew (the question was which of two events constituted the proximate cause of the claimant’s loss); The Maira (No 2) [1985] 1 Lloyd’s Rep 300 (the issue was the proper construction of a contractual document, which fell to be decided against the background of extrinsic facts which reasonable persons would have in mind in the situation of the parties and which shed light on the commercial purpose of the transaction); and The Wenjiang (No 2) [1983] 1 Lloyd’s Rep 400 at 402 and 408 (the issue was frustration which, although a question of law on ultimate analysis, had a large factual content, and in the answering of which the judgments and practices of commercial men played a large part).

33.  Mr Clayton argued to the contrary.  He submitted that the judge was not faced with an issue of fact, and that the grounds of appeal simply concern a question of law as to the correct interpretation of the words used in the MRs.

34.  In arriving at a proper interpretation of the MRs, a balance must be struck between the indications given by the language and the implications of competing constructions, which would include commercial common sense and checking the suggested interpretation against the commercial consequences.  The court should have proper regard to the views of the Arbitrator in respect of matters of commercial common sense and commercial consequences.  But overall, I am inclined to think that the grounds of appeal do not have a significant factual content.

Ground 1

35.  Bauer seeks to support the Arbitrator’s decision that the wording of the MRs was wrong in two specific respects, namely (a) the three horizontal dashed lines with an arrow in Diagram C1 of M12A marked “Founding levels determined by Engineer” should correctly say “Bedrock level”; and (b) the word “founding” in the phrase “theoretical founding levels” in the second paragraphs of M12D should read “bedrock” so that “Theoretical bedrock lines” as defined therein should be described as “the straight lines joining the successive theoretical bedrock levels”.

36.  The arguments of Bauer on ground 1 may be summarised as follows:

(1) The primary purpose of the MRs is the measurement of TITR. The MRs are not concerned with the engineering founding criteria which are set out primarily in the Sub-contract drawings, and which are separate and distinct.  Whilst the Method of Measurement is to some extent artificial, the objects of the Method of Measurement include at CESMM3 para 2.5^[35], which is intended to measure the work in a way which allows different classes of work which may give rise to different considerations of cost to be distinguished^[36].

(2) The starting point of the construction of Diagram C1 in M12A is the drillholes denoted by black vertical lines from which the Engineer determines the “founding levels” of D-walls. “Founding levels determined by Engineer” are signified by a dashed line with an arrow.  The connection of the “founding levels” at each drillhole as denoted by a small dotted line is called the “theoretical bedrock line” (referred to earlier as TB Line).  The TB Line is “theoretical” as there is no guarantee that the bedrock between the drillholes will have a “straight line” profile.

(3) M12A makes no mention of “theoretical founding levels” (referred to earlier as TFL). The only reference to TFL is in M12D 2^nd paragraph.  The MRs provide no reference, description or mechanism as to how or on what basis TFL is to be determined for the purpose of M12D.  There is no such concept as TFL.

(4) Diagram C1 must be construed in light of the true meaning of M12D.  The correct meaning of the 1^st paragraph of M12D is that once the results of the drillholes are known, the top level of the bedrock at each drillhole can be determined and agreed between the Engineer and the parties and the TB Lines can be drawn.  The Arbitrator correctly held that “theoretical founding levels” in M12D should read “theoretical bedrock levels” (referred to earlier as TB Levels) so that the TB Lines as defined shall be “the straight lines joining the successive theoretical bedrock levels”.

(5) By their nature and the determination by the Engineer, the “founding levels” referred to in the legend of Diagram C1 are the actual level at which the bedrock commences.  The dashed lines with an arrow indicate the top of the bedrock determined by the Engineer after the drillholes were available, not the founding levels which are typically 300 mm below the rockhead level for the toe in.  The JV’s interpretation whereby the TB Line for the purpose of the MRs is lower than the TB Line derived from the bedrock levels in the drillholes that were agreed between the parties based on the engineering requirements is absurd and contrary to how the engineering founding criteria were applied in practice.

(6) From the TB Line, the Engineer will determine the toe level of the panels in accordance with the engineering founding criteria found on the drawings (but not found in the MRs), see Drawing No. 823B/T/000/OAP/C01/004B.  In §1 of the Notes on Pre-drilling of this drawing, it is provided that for permanent D-wall, drillholes shall be sunk for every load bearing D-wall panel in order to establish “the diaphragm wall founding levels” based on the specified founding criteria. Unfortunately and confusingly, the toe level of the panels is defined on the drawings as “founding level”.  The “founding level” on the drawings is the toe level in the MRs, ie the bottom horizontal line in each panel in Diagram C1.

(7) It is manifest that the “founding levels” in the MRs are the actual bedrock levels determined by the Engineer from the drillholes from which the required toe level is determined.  It necessarily follows that the “bedrock level” and the founding level of a D-wall panel which is toed into rock cannot and will not be the same, nor can they coincide as shown on Diagram C1^[37].  The Arbitrator is manifestly correct to conclude that something had gone wrong with the language because of a confusion between bedrock levels and founding levels. The judge’s failure to follow the Arbitrator’s reasoning has resulted in an absurdity in that the “Founding levels determined by Engineer” in Diagram C1 are the same as or synonymous with the TB Line, when it is plain and obvious that such is not the case for a D-wall panel which is toed into bedrock.

(8) The judge has failed to give proper weight to the finding of the Arbitrator, who has substantial engineering experience and whose finding was made after hearing all the relevant factual and technical evidence on a point of interpretation with technical content.

37.  I do not think it could be said that the judge did not have regard to the analysis of the Arbitrator.  He acknowledged that his initial analysis agreed with the Arbitrator but on further analysis was not persuaded that the clear wording of the MRs would permit any rewriting of the contractual provisions in this regard^[38]. The Arbitrator’s construction involves assigning different meanings to important and defined terms in Diagram C1 and M12D to give the Rules a different meaning and effect.  The question is whether this rewriting of the contractual provisions to arrive at an interpretation which the Arbitrator considered would give a fairer and more commercially reasonable outcome is warranted in light of the language used.

38.  For TITR, the measurement formula given in M12B is as follows:

V (i.e. volume) = W x (P x D)

Where:

W = nominal width of panel as stated in the item description

P = length of each individual panel along its axis

D = depth measured from measurement horizon (referred to earlier as MH) to toe level of the panel

39.  The dispute is how “D” in the formula is determined.  For instance, if the D-wall panel is not to be keyed into bedrock, or for a panel which is keyed into bedrock but there is no depth between the MH and the constructed toe level, “V” will equal 0 and there is no TITR measured for that D-wall panel.

40.  For panels that are not to be keyed into bedrock, as mentioned earlier, M12F applies and this stipulates that the actual toe level of the panel shall be deemed the MH for measurement purposes.  To recap, the founding criteria for the Sub-contract (in Drawing No. 823B/T/000/OAP/C01/004B, under the heading “Permanent Diaphragm Wall Founding Criteria”^[39]) show where founding criterion (i) or (ii) is met, the panel is to be keyed into the category 1(c) rockhead.  Where the category 1(c) rockhead is not met for 15 m below the base slab bottom level (ie founding criterion (iii)), the panel can be founded on category 1(d) rock and is not required to be keyed into that material^[40], and is dealt with under M12F.

41.  For panels that are keyed into rockhead where founding criterion (i) or (ii) is met^[41], M12D provides that the theoretical bedrock level (referred to earlier as TB Level) at each drillhole location will be “the top level of the measurable core” (referred to earlier as MC).  The theoretical founding level (referred to earlier as TFL) is worked out based on the TB Level and the engineering requirements.  As submitted by Mr Clayton, the founding levels are “theoretical” because they are determined from the drillhole cores, and that the actual founding levels as further verified by the Engineer^[42] are likely to be lower, for the reasons as explained in the Judgment^[43].  I reject the contention of Bauer that there is no such concept as TFL.  I agree also with the judge that although the legend for the dashed lines with an arrow in Diagram C1 (“Founding levels determined by Engineer”) does not use the word “theoretical”, there is no reason to doubt that it refers to the TFL^[44].

42.  The TB Lines, defined in M12D as “the straight lines joining the successive theoretical founding levels”, will be determined in accordance with whether founding criterion (i) or (ii) applied at each drillhole location.  As the judge has observed, the drafter of the MRs must have in mind a distinction between TB Line and TB Level when these terms are used in M12D because different functions are to be served by them as explained above.  Nor is there reason to think that use of the words “theoretical founding levels” in the 2^nd paragraph of M12D was accidental, because “founding level” appear again in the 3^rd paragraph in that “Bedrock” is defined as “rock below the founding level as determined by the Engineer in accordance with the M&W Specification”^[45].

43.  M12E provides that for panels that are keyed into rockhead, the MH for each panel “shall be deemed to be a horizontal line sited 1.5 m above the highest point on the [TB line] within that panel”.  This allows “D” in the formula for TITR to be determined, as shown in Diagram C1, where the level of the MH is above the actual panel founding level for all three panels in the diagram, and the bottom horizontal line in each panel gives the actual founding level being the level to which the panel is toed in. The excavation measured from the MH to the toe level would be paid as TITR.  The “toe level” referred to in “D” in the formula has to be the actual founding level, referred to in the drawings as the “founding level”, as pointed out by Bauer.  There is no confusion when this is properly understood.

44.  I agree with Mr Clayton that the TFL referred to in M12D could not be the actual level the bedrock commences, as contended by Mr Pennicott, as the opening words of M12D are “Where diaphragm walls are to be keyed into bedrock”.  As mentioned earlier, M12D is dealing with the situation where founding criterion (i) or (ii) is met.  For founding criterion (i), the bedrock level as determined from the material in the drillhole cores might be well above the TFL.  Further, the TB Level at each drillhole location is defined in M12D as “the top level of the measurable core”, so the TFL could not be the bedrock level.  For panels keyed into bedrock, the use of the TB Lines to mean the lines joining the successive TFLs at each drillhole location will mean that the TB Lines would be above the actual founding level.

45.  Where the TB Line and the TFL are shown at the same point in the drillholes in the illustrations in Diagram C1, I do not think this should pose any difficulty.  I do not agree in light of this situation the definition of the TB Lines in M12D should be changed from “the straight lines joining the successive [TFL]” and rewritten as “the straight lines joining the successive [TB Levels]”.  To do so would be to ignore the clear definition of TB Level in M12D.  As mentioned earlier, the TB Line and TB Level serve different functions.

46.  For all the above reasons, I agree with the judge that effect should be given to the clear language of the provisions in M12A and M12D and there is no justification to rewrite the provisions in the way as the Arbitrator has done.  I would dismiss Bauer’s appeal in respect of ground 1.

Ground 2

47.  The question here is whether the definition of “measureable core” (MC as referred to earlier) in M12C, which imports the additional requirement of an RQD of at least 30, should apply to the same term MC as used in M12D, or whether MC as defined in M12C should only apply to the classification of Rock Class A and Rock Class B, as held by the Arbitrator and upheld by the judge.

48.  I will first look at the language of the relevant provisions before I turn to the commercial implications and matters of practicality considered by the Arbitrator.

49.  The starting point is M12B which sets out the definitions for the four types of excavated material: Other Material, Rock Class A, Rock Class B and TITR.  A formula was given for each type as to how “V” (ie volume) is arrived at.  I have set out earlier the formula for TITR. It is pertinent to note that in respect of the formula for each type, the measurement horizon (MH as mentioned earlier) forms part of the formula. As mentioned earlier, the MH is the central feature of the scheme for measurement and payment under the MRs.

50.  How the MH is determined is provided in M12D and M12E and is illustrated in Diagram C1^[46]. Where D-wall panels are to be keyed into bedrock, the top level of the MC at each drillhole location is the TB Level.  The TFL is worked out based on the TB Level and the engineering requirements.  Successive TFLs at each drillhole location are joined in a straight line to form the TB Line.  The MH for each panel is sited 1.5 m above the highest point on the TB Line within that panel.  It is clear that the top level of the MC is to be used in determining the MH under this scheme.

51.  If the MC is not given a consistent meaning throughout by applying the definition in M12C, the MH in the formula for other types of excavated material (Other Material, Rock Class A and Class B) would not be the same as the MH for TITR.  This would make the whole scheme unworkable.  It is pertinent to note that in the formula for Rock Class A and Class B, the value “K” is not just based on the length of the MC, but the length of the MC “above [MH] from drillhole(s) within the panel”.

52.  It is correct that in the definition for Rock Class A and Rock Class B in M12B, where reference is made to MC, it is expressly stated that the MC is “as described in Measurement Rule M12C”.  In respect of the reference to the MC in M12D, there is no such express stipulation.  Nevertheless, the same term MC is used in M12D, and it is only in M12C is there a definition of MC in the MRs.  Even though it is not expressly stated here that MC is as defined in M12C, I find it difficult to accept that a different meaning should be given to MC in M12D, in the absence of cogent reasons to the contrary. 

53.  As stated in M12C, the definition in this provision is “For the purposes of classifying rock to be excavated in [D-wall] trenches”. This brings us to the question if M12D relates to classifying “rock” to be excavated in respect of the TITR.

54.  Mr Clayton made these submissions in support of his contention that M12D does relate to classifying rock to be excavated in D-wall trenches:

(1) As stated in CESMM3 paragraph 5.5, “where excavation … is included in the work a definition of rock shall be given in the Preamble and this definition shall be used for the purposes of measurement”^[47]. Item 8 of Part C of the Preamble gives the definition of rock for different purposes, as required by CESMM3 paragraph 5.5.  For D-wall excavation, under item 8.1(b), “rock” is defined in the Preamble Part A Class C: Geotechnical and other specialist processes.

(2) Under item C2(a) of the Preamble Part A Class C, which relates to D-walls, TITR is stated to be one of the three classes of “Excavation in rock”, the other two being Rock Class A and Rock Class B.  Under item C2(b), which contained the MRs, TITR is defined in M12B as “that rock between the [MH] and the toe level of the [D-wall] as determined or instructed by the Engineer”.

(3) M12C is the only relevant definition of “rock” in this part of the Preamble. TITR is rock to be excavated in D-wall trenches, so the definition in M12C is to be used.

55.  Mr Clayton further submitted that it is illogical to have the requirement of RQD ≥ 30 for only two of the three classes of “Excavation in rock”.  It makes no sense for the requirement of RQD ≥ 30 to apply to Rock Class A and Rock Class B, which by definition is above the MH^[48] and hence on a higher level, but not to the lower excavation of rock in D-wall trenches for TITR, as generally rock quality is higher or less decomposed the deeper it is located.  He pointed out that the drillhole records included RQD values.  Further, the explanation below the illustration in Diagram C1 stated that the drillholes shown are used to determine “the founding levels of [D-walls] and lengths of “measurable core” for rock measurement”.  So MC as defined in M12C is to be used for “rock measurement”, not only of Rock Class A and Class B.

56.  On the wording of the contractual provisions, it seems to me that the language is sufficiently clear to support the interpretation that the definition of MC in M12C should apply to M12D.  The judge however took the view that it is not altogether clear if M12C is meant to apply to all rock^[49]. I turn to consider his reasons and Mr Pennicott’s submissions supporting those reasons.

57.  Even though the definition of TITR refers to “that rock”, the judge did not think M12C can be applied to “that rock” because TITR is a mixture of material.  He took the view that the terminology adopted in the MRs – ie referring to the two types of Class A and Class B rock as “Rock Class A” and “Rock Class B” – is to accord special treatment to these two classes of material, whereas TITR is a different category of material.  He saw considerable force in the Arbitrator’s analysis that M12C was designed to classify the rock to be excavated as Class A and Class B^[50].

58.  I agree with Mr Clayton that each of Rock Class A, Rock Class B and TITR is likely to have a mixture of material in the panel excavation for which payment is to be made for these categories, because the measurement of such material for payment purpose is based on the material in the drillhole only (which covered less than 0.1% of the area of a panel at any given level), and the material in the panel excavation is likely to be variable.  These considerations have been noted earlier in the Judgment^[51]. I do not attach any special significance to the terminology that the two types of Class A and Class B rock are described as “Rock Class A” and “Rock Class B”, as all three classes are under “Excavation in rock”.  The inconsistent use of inverted commas for various terms in the MRs^[52] is also not of significance.

59.  Mr Clayton had argued before the judge that M12D is not workable without M12C^[53]. M12D must be read with M12C so that category 1(c) rock with a total core recovery of at least 85% and fitting the founding criteria^[54] is the rock and bedrock referred to in M12D.  Otherwise, the definition of rock and bedrock in the M&W Specification^[55] (referred to in M12D) means that only Grades I to III material would be bedrock.  As category 1(c) rock is by its own definition recovery of 85% of such material, achieving the founding criteria with that definition would unlikely give rise to any TITR at all, because there would be around 15% of material which was not Grades I to III.

60.  The judge rejected the argument that M12D cannot work without reading with M12C.  To the contrary, he took the view that M12C is unworkable without the definitions of Rock Class A and Rock Class B in M12B (ie the length of MC for Rock Class A is more than 200 mm but not exceeding 1,300 mm and for Rock Class B is exceeding 1,300 mm), because without specifying the lengths of the MC, it is meaningless to speak of a total core recovery of at least 85% in M12C^[56]. The judge regarded this as clear indication that the definition of MC in M12C should apply only to Rock Class A and Class B.

61.  Mr Clayton submitted that according to “D” in the formula for TITR in M12B, the length of the rock for TITR, which has to satisfy the requirement of at least 85% recovery and RQD ≥ 30 on his construction, is from the MH to the toe level of the panel.  This is to be determined from the drillholes for TITR, as it is for Rock Class A and Class B, so there should be no difficulty in applying the definition of MC in M12C to TITR, as the percentage core recovery can be ascertained in any length of core. I think he is correct on this. I accept his submission that M12D is not workable without M12C.  The judge is wrong to hold that the requirements for the MC in M12D can be found in the engineering criteria for founding level^[57]. The engineering criteria contain no reference to MC and are for engineering purpose, not measurement purpose.

62.  I turn to the practical and commercial implications.

63.  The additional requirement of RDQ ≥ 30 in M12C is not found anywhere in the Sub-contract, save in the MRs, which were created to measure the excavation work.  This raises the question if the RQD requirement should concern the work carried out by Bauer.

64.  The judge took the view that the RQD requirement has nothing to do with the work performed by Bauer and there is no contractual basis to apply it to such work^[58]. He accepted the point made by Mr Pennicott there is a serious problem with importing the requirement of RDQ ≥ 30 into the excavation for panels which are to be keyed into rock, as the engineering criteria for such panels require the drillholes to extend at least 5 m into Grade III or better rock with a total core recovery of not less than 85%, ie category 1(c) rock^[59], and there is no equivalent guidance to identify the depth of the RQD requirement^[60].

65.  The judge noted that what took place was that the JV applied the Notes on Pre-Drilling in Drawing No. 823B/T/000/OAP/C01/004B so that the pre-drill was required to be at least 5 m into RDQ ≥ 30 material to satisfy the requirement for toe in (this is disputed by Mr Clayton).  There were many instances where there was no continuous core of RDQ ≥ 30. For these cases, the 5 m of RDQ ≥ 30 was relaxed and assumptions of where the RQD requirement was satisfied were made.  Consequently, there were cases where the RQD line and even the MH were below the actual toe level of the D-wall^[61].

66.  I do not attach special importance to the fact that the requirement of RDQ ≥ 30 in M12C is not found anywhere in the Sub-contract except in the MRs.  RQD is one of a number of potential cost differentiators, relevant to the ease and cost of rock excavation.  RQD does concern the work carried out by Bauer because rock with higher RQD values is generally higher quality with fewer natural features which affect the ease of excavation. There is no dispute that this requirement does apply to the measurement of Rock Class A and Class B.  This requirement has also been included for other excavation in rock at item 8.1(c) of the Preamble Part C, relating to ground investigation and bored cast in place concrete piles^[62].  As mentioned earlier, the drillhole records included RQD values.

67.  The MRs are rules created for the purpose of measurement and payment, with the measurement based on information available from drillholes, pre-drilled before the excavation of the panels.  Contractors and sub-contractors would tender on the basis of the applicable contractual provisions and the MRs served in part the purpose of allowing the contractor to tender for the works without full knowledge of the underground condition^[63], as drillholes were made after the Sub-contract was executed and data from drillholes were not available at the time of the contract.  In the absence of clear language to the contrary, it would be odd for the RQD requirement to apply to two items of rock excavation for D-wall panels, but not to the third being TITR.

68.  As to the practical difficulty envisaged by Mr Pennicott in applying the RQD requirement to TITR, I think Mr Clayton is right to say that the measurement could be undertaken on the JV’s interpretation and it is just the result of the measurement that objection is taken.  Applying that interpretation to the drillhole records, these records contain information whether there was a continuous core of RDQ ≥ 30 above the level to which the D-wall panels were excavated and this would show to what level there is the MC, ie material with a total core recovery of at least 85% of Grade III or better rock and RDQ ≥ 30.  As for the Notes on Pre-Drilling in the drawing, they addressed the engineering criteria for D-walls and the requirement is for drillholes to extend at least 5 m into Grade III or better rock with total core recovery not less than 85%.  It has not been the JV’s case in the arbitration that the pre-drill was required to be at least 5 m into RDQ ≥ 30 material and hence no question of relaxing such requirement.

69.  The Arbitrator took the view that if the RQD requirement is imported for TITR, the measurements are unlikely to be commercially reasonable. He referred to the examples of two panels, P071 and P192. These panels were toed into rock in a manner which satisfied the engineering criteria but were not paid at TITR rate.  Instead, they were paid at the rates for Other Material, Rock Class A and Rock Class B.  For panel P071, more payment was due (because of the higher volume of Rock Class A and Class B) and for panel P192, there was less payment. This led the Arbitrator to think that something must have gone wrong with the JV’s interpretation.  The judge agreed with the Arbitrator it is unlikely that the parties would have contemplated that panels toed into rock were not going to be paid at TITR rate and such a construction is not consonant with commercial sense^[64]. He opined that the inconsistency inherent in the JV’s construction means that it would be quite difficult, if not impossible, for a reasonable contractor to price for this project and to calculate the risk involved^[65].

70.  Instances in which panels were toed into rock in a manner which satisfied the engineering criteria but would not be paid at TITR rate or the results obtained from information in the drillhole records are not information reasonably available to both parties at the time of the contract.  Besides, commercial common sense is not to be invoked retrospectively. Even if the examples were somehow relevant, that must be weighed against the clear language of the relevant provisions.  As noted by the judge^[66], it is commonly understood that the measurement exercise pursuant to the MRs was never meant to be precise but a rough and ready measurement, and it is inherent in the methodology and the nature of excavation work that there is a risk the underground condition may turn out to be less favourable than assessed and vice versa.

71.  In my judgment, the judge has failed to follow the clear language and defined term in the contractual provisions and is obviously wrong to apply commercial considerations to rewrite the contract.  I would allow the JV’s appeal in respect of ground 2.

Ground 3

72.  The last question is whether “Bedrock” in M12D 3^rd paragraph should include category 1(d) rock.

73.  To recap, the Arbitrator answered this in the affirmative.  As M12D provides for D-walls to be “keyed into bedrock” and bedrock includes category 1(d) rock, when category 1(d) rock is less than 15 m below the base slab bottom level and at a higher level to category 1(c) rockhead, the D-wall panel will necessarily be keyed into bedrock and TITR is to be measured as part of the D-wall excavation if the D-wall is keyed into either category 1(d) or category 1(c) rock.

74.  The judge took a contrary view and held that category 1(d) rock cannot satisfy the definition of bedrock for the purpose of M12D^[67]. According to the definition in the M&W Specification^[68] (referred to in M12D), material containing Grades IV to VI cannot be bedrock.  The engineering criteria for category 1(d) rock in Table 2.1 of the Code of Practice for Foundations require a 50% minimum of core recovery of Grades I to III^[69]. If the core recovery of Grades I to III gets to 85%, it is category 1(c) rock, so category 1(d) rock may contain 50 to 84% of Grades I to III. As to the remainder^[70], it would be highly unlikely for there not to be any Grades IV to VI.  Hence, category 1(d) rock cannot satisfy the definition of bedrock.  So for panels that are to be keyed into bedrock, they must be keyed into category 1(c) and not category 1(d) rock.  Category 1(d) rock is only relevant under founding criterion (iii)^[71], and in those cases the panel needs only be founded on category 1(d) or better rock. 

75.  Mr Pennicott submitted that the Arbitrator is correct and adopted his reasoning. The main arguments advanced may be summarised as follows:

(1) According to the definition in Table 2.1 of the Code of Practice for Foundations Category 1(d) material is rock, being “moderately decomposed, moderately strong to moderately weak rock of weathering material grade better than IV, with a total core recovery of more than 50% of the grade” (emphasis supplied). In Table 1 the Founding Material Schedule^[72] of Drawing No. 823B/T/000/OAP/C01/004B, under the column “Rock material decomposition grade” (emphasis supplied), there are two types of material: category 1(c) or better, and category 1(d).  By this categorisation, again category 1(d) is rock.

(2) As the definition requires category 1(d) material to be “rock of weathering material grade better than IV”, this means it has to be Grade III.  With reference to the definition of bedrock in the M&W Specification (“Bedrock” means rock which is defined by description to have been an originally solid in place rock mass below which Grade IV, V or VI material normally would not or has not been found), category 1(d) is bedrock.

(3) The reference to “core recovery” in the provisions mentioned above means that for category 1(d), the fact that it is better than Grade IV must be established by a core recovery of more than 50% of the material and is a measure of the quality of the core.  It does not mean that a percentage of the material will be Grade IV^[73]. It is not permissible for the judge to speculate it would be highly unlikely for there not to be any Grades IV to VI material in category 1(d).  By parity of reasoning, on the judge’s analysis, if the amount of Grade III gets to 85% but all or part of the remaining 15% is Grade IV, then category 1(c) cannot be bedrock either.  Only category 1(a) rock would classify as bedrock since this is the only category of rock with 100% total core recovery.

76.  It is clear from the founding criteria in Drawing No. 823B/T/000/OAP/C01/004B that D-wall panels would only have to be keyed into category 1(c) rock and it is category 1(c) rock head that is relevant, see founding criteria (i) and (ii).  For founding criterion 1(iii), where the category 1(c) rock head is more than 15 m below the base slab bottom level, category 1(d) rock is not expressed as a “rock head” in this founding criterion, which only requires the panel to be founded on category 1(d) or better rock.

77.  The Arbitrator sought to address this by saying that there would be no keying into category 1(d) rock, and M12D would not apply and instead M12F would apply, when the category 1(d) material is below 15 m and forms a horizontal surface at the founding level^[74].  But I am inclined to agree with Mr Clayton this would mean that effectively no panels would fall within M12F, as horizontal is precise and the rock profile is variable at any given level^[75].

78.  The relevant word in the present context is bedrock rather than rock.  It is clear from the definitions of rock as quoted above, a total core recovery of more than 50% of the grade means that for category 1(d) rock there only needs to be more than 50% of better than Grade IV material. According to the definition of bedrock in the M&W Specification, bedrock is rock mass below which Grade IV, V or VI material normally would not or has not been found.  The Arbitrator’s holding did not address the question if category 1(d) rock were regarded as “bedrock” for the purpose of M12D, and if the remaining 49% could not be Grade IV, V or VI material, what material could that be.

79.  As for the point about “core recovery” and that category 1(c) cannot be bedrock either on the judge’s analysis, I have accepted Mr Clayton’s earlier submission in ground 1 there is no difficulty when M12C and M12D are read together, as under M12C there has to be 85% recovery so that category 1(c) rock, which fits the founding criteria, is the rock and bedrock referred to in M12D.

80.  For the above reasons, I agree with the judge that “Bedrock” in M12D should not include category 1(d) rock.  I would dismiss Bauer’s appeal on ground 3.

Conclusion and costs

81.  I would allow the JV’s appeal in CACV 113/2019 and dismiss Bauer’s appeal in CACV 122/2019.

82.  As the JV is successful on ground 2, the judge’s order depriving it of half of the costs should be set aside and substituted with an order that the JV should have the whole of the costs of the appeal before the judge.

83.  For the costs of the appeal, we have heard the parties on costs and there is no dispute that costs should follow the event.  I would award the costs of both appeals to the JV.

Hon Cheung JA:

84.  I agree with the judgment of Kwan VP.

Hon Barma JA:

85.  I agree with the judgment of Kwan VP.

Annex 1

Annex 2

Annex II

ATTACHMENT C

Figure 1 – Example of D-Walls and Base Slab

Figure 2 – Founding Criteria

Annex 3