Website Banner. John Monash: Engineering enterprise prior to World War 1.

[Main Index] [Projects Overview] [Bridges] [Buildings] [Tanks] [Walls] [Wharves etc] [People] [Localities] [Abbreviations] [Units & Currency] [Glossary]

Marine and Riverine Projects:
Wharf for John Darling & Co., Birkenhead (Adelaide).
Unsuccessful tender.

Summary.

This encounter between Monash and C L Gardiner, a consulting engineer committed to timber construction, raised important questions in the assessment and handling of risk. Gardiner had prepared a design for a timber wharf at Birkenhead, but invited the South Australian Reinforced Concrete Co to prepare an alternative design in that material. Strangely, Gardiner did not specify the amount of superimposed load that the wharf would be expected to carry, so Monash worked back from Gardiner's proposed beam sizes and decided they were intended for a load of 200 pounds per square foot of deck (200 psf = 9.5 kPa). He designed his concrete wharf for this value using his usual factor of safety of 4. The estimated price proved higher than that for the timber wharf, but competitive if its lower maintenance costs were taken into consideration.

Gardiner rejected the SARC proposal out of hand. He declared that the wharf was required to carry 560 psf, and claimed that his timber design would be able to withstand twice that. After arguing questions of safety, SARC reluctantly prepared a design for a concrete wharf to carry 560 psf, but with a reduced safety factor of 3. However, the price was now quite uncompetitive and SARC's final proposal was ignored.

At the time, the 'permissible stress' concept was used for ensuring a factor of safety in structures. Greatly oversimplified, the idea is that if tests show that a material will fail at a stress of, say, 20 units, the designer should choose the dimensions of components to ensure that the stresses caused by the expected loads will not exceed, say 5 units. With such figures, the factor of safety would be 20/5 = 4, the value used by Monash.

There is nothing in the RCMPC records to show what calculations Gardiner may have made, or whether he relied simply on traditional practice. He may have used a factor of safety less than 4, and/or assumed a much higher strength for the jarrah timber than Monash was prepared to allow. It seems that the only argument he put forward was that the dimensions of his wharf would be identical to those of existing timber wharves and that these had performed successfully under 560 psf. Had Gardiner selected very high-grade timber for these wharves, or had their owners been unknowingly working them dangerously close to their limits? Or was Gardiner simply outwitting his commercial rivals?

The Story.

Invitation to tender.

On 3 February 1911, H G Jenkinson, Resident Engineer for the South Australian Reinforced Concrete Company (SARC) informed Monash that John Darling & Sons were calling tenders for a timber wharf on the Port Adelaide River. "Mr C L Gardiner of the Canal Company, who is the Engineer to the above firm, states that it would be well worth our while to submit a design and estimate for r.c. construction". Jenkinson enclosed a drawing and specification, but noted that Darling & Sons had only a 21-year lease on the site. (This would detract from Monash's main selling points: that reinforced concrete was "practically maintenance-free" and "everlasting", and thus well worth its higher initial cost.) SARC General Manager E H Bakewell had cautioned: "John Darling is a very keen business man and we must have nothing but hard and fast agreements". Tenders closed on 9 February.

General note. The location of the wharf was described as being on the left bank of the Port Adelaide River, stretching "from Elder Road to Birkenhead Street". The latter is now Victoria Street. Both it and Elder Road ran in those days in a straight line to the river bank. This would place the wharf just downstream from the Birkenhead Bridge.

Designs and Cost Estimates.

Monash received Jenkinson's letter on 4th and immediately set to work. Using Gardiner's drawing, he estimated the likely cost of the timber version as £4642. His reinforced concrete structure would have to compete with this. Working backwards from the dimensions shown for the timber bearers supporting the deck, and assuming a permissible bending stress for jarrah of 800 pounds force per square inch (800 psi = 5.5 MPa), he deduced that Gardiner had adopted a design live load of 200 pounds per square foot (9.5 kPa). Monash prepared his rough design in reinforced concrete to the same criterion. With substructure and piles included, he arrived at a total cost of £3871. To this he added £387 for contingencies, administration, and sundries, and £852 representing a 20 per cent margin - though he thought this somewhat on the low side. This brought the total estimated price to £5110.

Technical note. Although JM designed the main deck beams for an overall load of 200 psf, he followed standard practice in designing the smaller secondary beams for 220 psf and the deck plate itself for 250 psf, to allow for possible higher concentrations of load over small areas.

On 6th Monash wired Jenkinson: "Try for extension tender time. Very busy. Am short handed here." Both his senior Assistant Engineer, P T Fairway, and RCMPC's Managing Director John Gibson were on holiday. Monash was obliged to keep an eye on business and accounting matters as well as engineering. He therefore sent to Adelaide what he had achieved so far: a blueprint and specification, a file of calculations, and a pencil sketch of details.

He told Jenkinson: "Neither arithmetic, design nor sketches have been in any way checked, nor am I sure that my several assumptions as to surrounding conditions are correct. Therefore, use my papers as a guide only, and review every aspect … I have had no time for any revision; so better read them through entirely before drawing any inferences." By careful re-design, it might be possible to reduce the bulk and cost of many components. Gardiner's timber wharf would be "a jimcrack affair", worth only about 1½ cwt per square foot (168 psf) superimposed load. If Darling & Sons would confirm that this was indeed all they expected, Jenkinson would be able to cut down heavily on the strength [and therefore cost] of the reinforced concrete version.

There was another way in which savings could be made. The ground behind the wharf had to be built up with earth fill to provide a working surface. Because Gardiner had doubted the stability of the existing ground which sloped towards the river, he planned to excavate down to a thin layer of limestone that occurred about 30 feet below deck level. Sheeting would then be required behind the rear piles, extending down to the limestone, and strong enough to retain the full depth of fill. Monash was confident that the natural ground was reasonably stable and that the sheeting and earth fill need go down only as far as the present ground surface (diagram below). "We should be able to evolve a design of equal strength for about £5100 to £5400. Suppose it costs them £800 more than timber. This divided over their 21 years lease will cost them less than £40 a year - which is trifling compared with the upkeep of their timber wharf - which will begin in a year or two and get heavier and heavier."

Because time required for construction was an important factor, SARC's proposal suffered the disadvantage that reinforced concrete piles could not be driven until they were 30 days old. However, Monash pointed out that the piles would be relatively short, and there should be no problems with seasonal weather.

Cross-section through SARC's proposal for a reinforced concrete wharf structure.
Based on a drawing in the J Thomas Collection.
HWL = high water level. LWL = low water level.

Technical note. The design incorporated the usual double row of piles along the river bank. Near the top, each front pile was connected to the corresponding rear pile by a reinforced concrete diaphragm wall carried on a beam. These assemblages were to be spaced every 20 feet along the river bank. The deck plate was to be supported by beams running parallel to the bank, spanning from diaphragm to diaphragm. The diaphragms were to be 11 feet deep, with a thickness of 9", increased to 12" near the front to take a vertical fender of 12" × 12" jarrah. To retain earth fill behind the line of rear piles, a series of reinforced concrete "slabs" was proposed, spanning horizontally between piles, and varying in thickness from 14 inches at natural surface level to 6 inches at deck level. Piles were to be 16" × 16". Front piles were to be driven to 50 feet below deck level and their tops cut down to bond with the cross-beam. Rear piles were to be driven to 40 feet below deck level, passing through a crust of limestone lying about 15 feet below Low Water Level. Their tops were to be cut off at deck level.

Jenkinson checked Monash's design and estimate, and added ancillary items such as fendering and mooring facilities, arriving at a basic cost of £4677. To this he added a margin of £1200, giving a tender price of £5877. He thought the price of the timber design would be at least £5000, but was not confident that Darling & Son would be willing to pay extra for the greater durability of concrete. Bakewell, SARC's Managing Director, thought £5877 was too low; but Jenkinson argued that it was only "by constantly quoting" that SARC would ever be able to break into the local wharf construction business. SARC's letter of quote on 11 February pointed out that the whole structure would form a "complete monolith" and be "everlasting". The sheeting would go down only so far "as deemed necessary". Completion was promised within 8 months.

Perceived strength and risk.

As must have been expected, Darling referred the quote to Gardiner, who challenged SARC to guarantee their design for 5 cwt/ft² (560 psf). Jenkinson appears to have taken no notice of this, as on 22nd he issued a working drawing for his scheme as it stood (200 psf). Gardiner called in personally to SARC's office and reaffirmed his position: SARC had failed to meet his specifications; his own design was superior.

Jenkinson told Monash that he did not doubt Mr Gardiner's word, "but the factor of safety in such cases must have been perilously low. Timber structures as we all know can be subjected to stresses far above those for which they were designed without showing signs of failure. In these circumstances it behoves us I think to consider the proposal to lessen our usual factor of safety in such cases as this. It is ridiculous for Mr Gardiner to make out that his Wharf … was designed for 5 cwt … added load". He hoped to convince Gardiner by showing him the solid dimensions proposed by government engineers in their design for a timber wharf at the Ocean Steamers' terminal, where the design load was indeed 5 cwt/ft².

Learning that SARC's tender of £5877 had been just below the lowest tender for the timber version, Jenkinson made one more bid. He at last re-designed the structure to meet Gardiner's specifications. Importantly, in order to keep the price down, he reduced the factor of safety from 4 to 3, arguing that "the only occasion when the wharf is likely to be loaded to this extent will be at the official tests". Even so, he was obliged to reduce the length of the 'bay' between diaphragms to 12 feet. The new cost worked out at £5801 to which was added £500 for contingencies, administration etc, and a margin of £1300, giving a total price of £7601.

Monash thought it was hardly worth pursuing the contract with a price £1500 more than the timber version, against the "subtle opposition" of Gardiner. The price was too high to compete and the margin too low. Jenkinson countered that the price was not the point. "Our new price will be pitted against those for Mr Gardiner's Timber design. This is not fair. Mr Gardiner has evidently told Mr Darling that his timber design was amply strong for a 5 cwt load. Are we to actively combat this?" He was supported by John Gibson who felt, as JM put it, that "a matter like this should be pushed for all it is worth until it is lost beyond recall".

So Monash advised Jenkinson to word his tender in such a way that the cost of the structure could be cut down afterwards, when it might be possible to replace the diaphragms, which represented half of the total superstructure concrete, with cross-bracing. Darling & Sons should be told how much they would be paying for the extra-strong deck and the deep back-fill, in the hope that they would dispense with them "even though they might have some academical doubts as to the efficiency of our method of retaining the back filling". If time permitted, Jenkinson should examine the cost of a larger bay spacing of 15 or 16 feet. This would increase the size of the longitudinal girders, but reduce the number of piles, anchorages and diaphragms and so might reduce the overall cost.

Although Jenkinson had earlier suggested that it would be unwise to discredit Gardiner with Darling, "for the two gentlemen are very friendly", Monash commented: "I say without hesitation that if there is no chance of getting the job on any other terms, I would not hesitate to tell Mr Darling straight that Mr Gardiner is quite wrong in his assertion that his wharf is quite strong enough to carry 5 cwt. per square foot. The contention is too absurd to permit of it being passed by without some protest".

The new quote therefore took the form of the basic figure of £5877, plus two 'optional extras': £800 for driving the sheet piles down to the limestone layer, and £900 for increasing the deck strength to a reliable 5 cwt per sq ft. Jenkinson's carefully-worded covering letter argued that timber structures designed on traditional lines had been pushed unknowingly far beyond safe limits, whereas the scientific methods developed for reinforced concrete established a clearly defined, numerical factor of safety.

 
The comparison of prices under these inequitable conditions cannot set off the reinforced concrete construction to much advantage from a financial point of view. We respectfully ask you to give these points due consideration before finally settling the matter. Should you decide on having a reinforced concrete structure you will obtain a wharf constructed on the most modern principles and superior in every way to one made of timber.
 

There the correspondence ended.

General Note. At the time of this project, the process of establishing the engineering properties of Australian timbers was at an early stage. Quality control and design of timber for engineering structures is now well developed and scientific. Concepts, techniques and magnitudes for safety factors are nowadays specified in published Standards and 'Codes of Practice' drawn up by committees. These include research engineers, designers and representatives of the construction industry: suppliers of raw materials, fabricators and builders. The Standards are referred to in laws governing building permits.