Holdfast Training Services Trial Blog

Australian SAS Operator

Introduction

In similar fashion to the comments on Riche’s latest CPD blog, I too have been hunting out more formalised CPD activity opportunities. Although there are some within the design office I wanted to broaden my search and so started looking for opportunities outside of the workplace. Namely seminars, conferences and workshops delivered by professionals in the building services sector and that are affiliated to various engineering institutions.

This blog briefly discusses a seminar I attended hosted by Lighting Options Australia as part of the Society of Building Services Engineers WA Chapter in association with CIBSE.

Lighting Options Presentation

The seminar consisted of the usual finger food and drinks reception, with a studio tour followed by a technical presentation on LEDs in 2016, which culminated in a Q & A session.

The speaker, co-founder and MD of Lighting Options Australia has been involved in lighting throughout his 17 year working career, combining on-the-job experience and knowledge whilst working in partnership with several internationally renowned lighting brands.

Before the presentation kicked-off the speaker demonstrated some of the lighting options used in their showroom gallery.  One painting of interest was the Australian SAS Operator, which was painted by an ex-serviceman’s wife.  It is said that whichever angle you look at it from, it always looks as though he is aiming directly at you.

The presentation covered three areas:

1. LED performance in 2016.
2. LED/Luminaire life and lumen maintenance.
3. Introduction to Richard Kelly’s ‘Language of Light’.

Presentation Summary

1. LED performance in 2016. Four factors were discussed: efficacy, colour tolerance, colour rendering index and failure rate.

• Efficacy. Luminous Efficacy is the measure of how well a light source produces visible light (to the naked eye). It is the lumen value against the energy consumption, measured in Lm/W.  Manufacturers usually display this technical data along with the colour temperature, measured in Kelvin (K). For example, 100 Lm/W (3000K) or 130 Lm/W (4000K). The colour temperature describes the impact of colour, which gives either a warm light (<4000K) or a cold light (>4000K).

• Colour Tolerance. This is the colour consistency amongst exact same light sources, for example, an array of lights along an art gallery wall, where a higher colour variance in one source could be very noticeable and detract from the required lighting effect.  There is a standard/tolerance for the acceptable degree of variation in colour temperature, ANSI specification C78.377-2008. The tolerances can be identified by applying the MacAdam Ellipses. The size of the ellipses is based on the standard variance called ‘steps’, the more steps the higher the variance and more obvious to the naked eye (see figure 1).

Figure 1. Colour Tolerance measured in MacAdam Steps

With the innovation of LEDs moving lighting technology away from the more traditional incandescent or compact fluorescent light sources, it introduces more variables where the Correlated Colour Temperature (CCT) of LEDs is liable to move further away from the target colour. Typically, incandescent and fluorescent ellipses don’t exceed 7 steps. Therefore, due to the easier identified colour variance, the LED standard states that CCTs must be within 4 steps MacAdam. The following link explains colour tolerance in more detail and is the source of reference.

http://www.lampslighting.co.uk/colour-tolerances/

• Colour Rendering Index. Colour rendering measures the light source’s ability to render colours correctly and is graded from 0-100. The old standard was to use eight colours to act as the controls against a pre-defined light source. This has now been increased to a 99 colour sample (see figure 2). The new sample range is more representative of real-world objects as opposed to the original and is intended to fairly and accurately characterise LED and legacy light sources.

Figure 2. Colour Rendering Index.

Failure Rate. When conducting lighting design and planning off failure rates it is important to understand that traditional, incandescent and fluorescent, light sources are failure rated at 50% over their rated life, which indicate that the lamp has failed to the point of needing replacement.  However, the standard failure rate in the LED market is 0.2% for 1,000hrs. That works out as 10% after 50,000hrs (the standard LED manufacturer specification failure rate which is extrapolated from an industry accepted standard of 8,000hrs burn time). The latest in LED technology, as manufactured by ERCO, has decreased this by 100 times to <0.1% for 50,000hrs. By example this means if using 100 x 10 LED chip luminaires in a project then only one single LED out of 1000 might fail after 50,000hrs.

Figure 3. ERCO LED Failure Rate

2. LED/Luminaire life and lumen maintenance.

This part of the presentation discussed the characteristics that determine LED life and maintenance. There are four values:

1. L – This is the light output of a LED module which decreases over its lifetime. L70 means the LED module will give 70% of its initial luminous flux. This value is always related to the number of operation testing hours (usually 8,000) and therefore is a statistical value so a batch of LED modules may have a slight variance in their lumen maintenance. A point to note is manufacturers should state the testing hours of their product as some try and mislead their customers by only testing to 4,000hrs. This means there is likely to be a higher percentage of error when extrapolating to the 50,000hrs rated life but they clearly do this as it halves their testing time to approx. 5.5 months thus cheaper to produce and get to market. If a product doesn’t say it’s testing hours it is safe to assume it is 4,000 and not 8,000hrs
2. B – This is the degradation value of LED modules which are below the specific L values. So, L70 B10 means 10% of the LED modules are below 70% of the initial luminous flux.
3. C – This is the value of fatal LED module failures, indicating the percentage of modules that will actually fail.
4. F – This is the combination of B and C values, for example, L70 F10 means 10% of the LED modules may fail or be below 70% of the initial luminous flux. As a rule of thumb the most commonly used values are L and B. The LED market standard specification currently used is L70 B50 (50,000hrs), e.g. after 50,000hrs only 50% of the LEDs used still achieve 70% of their original luminous flux. ERCO uses LEDs with the specification L80 B10 (50,000hrs), e.g. after 50,000hrs at least 90% of the LEDs still active achieve 80% of their original luminous flux. Figure 4 shows the two graphs for comparison.

Figure 4. Traditional Vs ERCO LED Specification Maintenance/Failure Rate

Here’s the difference in a real life example:

There are 10 x LED downlights used to achieve an average 500 lx and rated at L70 B50 (50,000hrs).

At 50,000hrs – 500 lx x 0.7 = 350 lx

At 50,000hrs – 350 lx x 0.5 (worst case) = 175 lx

The same 10 x LED downlights are used to achieve an average 500 lx but this time rated at L80 B10 (50,000hrs).

At 50,000hrs – 500 lx x 0.8 = 400 lx

At 50,000hrs – 400 lx x 0.9 (worst case) = 360 lx

So you can see using the L80 B10 specification retains 72% of the original lux level after 50,000hrs versus just 35% if using L70 B50.

3. Introduction to Richard Kelly’s ‘Language of Light’. The final part talked about the American lighting designer, Richard Kelly, who was said to have been one of the pioneers in architectural lighting design. One of his goals was to be able to get both architects and lighting design engineers to speak the same ‘language of light’. To achieve this he came up with a very basic concept that broke light down in to three simple elements:

1. Ambient – ‘Ambient luminescence’ is the element of light that provides general illumination, ensuring the surrounding space, its objects and any people in it are visible. This form of lighting facilitates general orientation and activity. Ambient luminescence is the foundation for a more comprehensive lighting design and aims to have differentiated lighting that builds upon base layers of ambient light.

Ambient

2. Accent – ‘Focal glow’ is the light that helps to convey information and guide movement. Brightly lit areas automatically draw our attention. Directed light accentuates focal points and helps to establish a hierarchy of perception using brightness and contrast, helping to emphasis important areas and accelerate spatial orientation.

Accent

3.  Scenic – ‘Play of brilliants’ results from the ability of light to represent information in and of itself. It covers a multitude of lighting effects used for their own sake, for atmospheric or decorative reasons, but having no specific practical function. Examples include, an emotive candle on a table, a fire place, an object of coloured light being used to influence the ‘climate’ of a space.

Scenic

Examples of all three elements combined can be seen in the following photos:

The State Theatre, New York

Grovsvenor Place, Sydney

171 Collins Street, Melbourne

Forthcoming Events

Having chatted to the CIBSE Fellow who organised the event I am now in his ‘in-tray’ for any upcoming events which he assured me are roughly one per month covering various building services topics.

Since arriving at KBR I have been involved in various elements of Safety in Design (SID).  I don’t recall the subject being touched on in great detail on Phase 1, but it forms part of the core business of the design team.  This blog gives at introduction to the systems used by KBR and some of the key issues I have witnessed so far.  The basic steps in the KBR SID process are covered below, how the fit into the design process is illustrated in Fig 1.

Hazard Identification (HAZID) studies are generally conducted at an early stage in the design by senior designer and the client. Identifying key hazards early means that they are revealed before significant costs have been occurred in the design process.  It also means that if they are unavoidable further risk reduction measures are actioned during the detailed design.

Construction Hazard Assessment Implication Review (CHAIR) is a tool to assist designers, constructors, clients and other key stakeholders to come together to reduce construction, maintenance and demolition safety risks associated with a design. The design can be considered as a whole or sub systems.  Generic guidewords are used to identify hazards associated with three stages of a project; concept, construction (also including demolition) and maintenance and repair.

Hazard and Operability (HAZOP) studies are carried out as the detailed design is beginning to go firm. It is a detailed hazard and operability problem identification process, carried out by a team generally including representation from the client, operators/maintainers, suppliers, designers and installers.  The design is broken down into ‘nodes’ related to how the system is designed to operate.  Then various guidewords are used to provide out of normal operation scenarios.  For each one the risk, cause and consequence are recorded, along with a solution if agreed or an action.

Control Hazard and Operability (CHAZOP) studies are carried out once the control narrative is complete.  Similar to the HAZOP the CHAZOP study uses a number of guidewords to study deviations from design intent caused by computer/control issues.

Although this looks like a very linear progression in reality it is very iterative with the feedback from various stages been integrated and reviewed before the next. All of these processes are recorded and form part of a SID report for each project.  This then provides a log of how a design developed and why decisions were made. It forms a key piece of evidence should something go wrong with a plant in the future.

Fig 1 – The Design and SID Process

So far I have facilitated a couple of HAZOP and CHAIR meetings already and taken part into a few more.  Some of the issues I have picked up so far include:

• The importance of the leader – the lead of the workshop is a key role.  They need to keep the group on track, ensure all key information is recorded and prevent discussion going down various potential rabbit holes.
• Time Management – The key to the meetings is to identify hazards not solve them.  Should a point take more than the allotted ten minutes it needs to be reviewed separately.
• Time Allocation – The workshop needs to have sufficient time allowance and should be broken up if possible.  Generally the sessions I have been to have lasted all day and the level of interest thus the quality of the output drops throughout the day.  Sequencing to allow the high risks elements first can also assist with this, as do sugary snacks.
• Design Quality – The design needs to be of sufficient quality and detail to allow the workshop to take place, otherwise it’s a waste of everyone’s time.
• Preparation – The designs need to be sent out in ample time for everyone to review and submit comments prior to the workshop – otherwise you risk the workshop disintegrating into a design review.
• Attendance – Make sure the correct number of people are present.  Too many and the discussions take too long, two few or the wrong attendance and key items could be missed.
• Variations – Often this is the first time the client has looked at the design in detail and from my experience so far they try to use the opportunity to add things to the design.  As mentioned previously the purpose of these workshops is to identify potential hazards, not work out who is commercially responsible.

Bank Street – If Carlsberg did risk mitigation for piling

I attended a site visit to Bank Street at Canary Wharf this week. Arup are the design engineers and Laing O’Rourke (Expanded Piling) are doing the piling. The activities are mainly piling so far there have been an incredible amount of problems, solutions and wonderings to be discussed. I’ve attempted to highlight some here.

The current site – note central CHS piles for reference later.

CFA piling

Test pile

The plan was to install the tension piles for the main test pile but the rig being used could not extract the auger once it had reached toe level. The issue was found to be that as the top of the auger went below ground level the material on top of it tended to pack it in on top so trying to extract resulted in the mass of soil above the top of the auger being too great to remove. Therefore a longer auger, one that extended the full pile length was needed, and an equivalently long mast. This was realised and a new rig was used with success.

As part of the specification each pile installation was checked for the rate penetration with auger rotation. The aim being to get 10 rotations per metre. This reduces any flighting where soil material is dragged into the pile and therefore reduces the pile diameter. This worked well and the rig can be automatic to achieve this.

Example of getting the flighting penetration rate too slow or too fast.

Source:  http://www.fps.org.uk/fps/guidance/safety/CFA%20Piling%20Preventing%20ground%20rig%20instability%20through%20over-flighting%20FINAL.pdf

Instrumentation calibration

This is not something I really picked up on while on site Two Fifty One but is really important, especially with CFA piling as one can’t look inside the casing to check the base.

Pressure gauge – CFA piling relies upon positive pressures when filling the pile as the auger is extracted. The positive pressure implies that the concrete is being distributed all around the bore to construct the pile. The rig on this site read 0.5bar so seemingly all was well. However, something was amiss – When it came to inserting cages they only went a few metres until it was found the pile had collapsed in. The strokes on concrete pump were counted to ensure the correct volume was being distributed into a 2metre cubed box. It wasn’t so a pressure gauge test was done on the rig sensor and it was found to be faulty. The pile was being under fed by about 20%!

After the many weeks of trying to get these piles sorted for the pile test, the rig (new one) and old rig instrumentation was mended. My advice, or what I would do now I know it, would be to ask to see some of this calibration going on if I were on site while piling is going on. Arup put these tests in their specifications to ensure everything is working which seems sensible.

Cased, rotary bored piles

Secant wall

CFA rig with Secant wall behind.

Ground anchors along length of secant wall. Note which piles the anchors are in (female…?)

The secant wall was due to carry both horizontal and vertical loads. It was base grouted which is where the stiffness below the pile is improved, not ultimate bearing resistance, to reduce initial settlement. This is where the pile cures and sometime later the steel tubing inserted inside the cage is pressurised with water which cracks the bottom of the pile. [I have a photo on the way, but for now imagine a steel hose pipe running down to the bottom of the cast pile with a couple of holes in the bottom.] Grout is then pressurised inside the tube and in theory fills any voids around the base. A subsidiary check is to get a pressure to show that the grout is not just heading off somewhere. There should also be some pile uplift to prove base resistance. This was not always achieved which in theory would be a concern. But it was a secant wall so if uplift was achieved, would the piles have actually have been connected? The stratigraphy at the site is such that it is not far from ground to Thanet sands so working loads can’t all be taken in the shaft. Therefore it’s more important to understand about base resistance. One might question the point of doing a test with secant piles, other than with the confidence of reaching a certain pressure.

As a quick tangent, my colleague who worked on Cross rail, where penetration grouting was used, advised on the need to avoid over pressurising the base of piles because of the risk of heave. They only had 7m of soil above the base of the pile, let’s say 140kPa. 1 bar is 100kPa. Apparently you need 4bar to get the grout into the voids. Mmm turns out there was some heave!

Back to the site visit.

Double skin sheet pile wall

This was where the most fun occurred. 2 rows of sheet piles were installed to act as a self-supporting wall on one edge of the site.

Near: Double sheet pile wall. This is filled with material then ties are placed and then the top section was concreted. Rear: Secant wall installed, reinforcement for capping beam can be seen.

Double wall cofferdam from the inside

The sheets were installed by pre-augering. Angela raised this as a question last year I think. Arup advised against it. As that was what the contractor wanted to do, Arup specified some come penetration tests either side of the piles to see what resistances were gained (minimum of 15 N value). A chart was presented down to the Thanet sands showing almost no resistance! The augering inside the piles had removed lots of soil and where it had influenced outside, it had reduced the soil strength to close to zero.

So what? The solution was to fill inside with concrete and grout the outside. Concrete was used to avoid having to go back and certify (if granular material was used it would probably have been difficult to achieve any compaction and may have failed further CPT tests, using concrete avoided that). This took months, was expensive and difficult to do underwater. Any plans to reuse the sheets anywhere below ground level are now out!

Summary

Firstly it was great to be back on site and see some piling elsewhere compared to Two Fifty One. The engineer had clearly been busy resolving problems but it seemed to be that most of the tests, checks and procedures were all sensible and go some way to mitigating what is really a pretty uncertain activity. One assumes all is well with the pile but really we have no idea. There is a balance between equipment, instrumentation and people. They kind of all need to chime together to reduce the risks. I feel that this visit was very useful confirmation of the sort of standards expected when piling in quite a range of issues.

More photos

Vibrating head with circular hollow section to push reinforcement cage to pile cut off level. We used a similar method at Two Fifty One, but without the vibrating head – this methods looks far easier but needs more equipment.

The site before dewatering. Left: double sided cofferdam, centre: CHS piles shown above for orientation, right secant wall.

On the way in we saw piling under bentonite. Rig with digging bucket attached removing spoil. Bentonite silos (black), green objects in front are the cleaning chambers to remove granular material residues.

Design process

Background

The method of getting sewerage from the current outfall point in the Thames into the new Thames Tideway sewer is via a series of chambers and a 30m shaft. At the Putney site these chambers are located on a foreshore site. To facilitate the construction a cofferdam is required. 

This blog looks at the approach taken to decide how to size up some piles for design development.

Putney Embankment Foreshore – End state river wall is the rectangular box in the foreground.

Durability

Tideway (the client) require a 120 year design life for all structures. This is onerous but after Guz’s recent blog on his river travels I can see why there is pressure to prove everything meets the design life. 

Tender stage

Tender documents had the permanent cofferdam as a secant wall. To enable construction an outer vast temporary sheet piled cofferdam was proposed.

The blue dotted line is the extent of the tender document sheet pile wall. The orange box is the proposed site location.

The contractor’s tender submission was an optimised design by combining cofferdams to have a single thicker sheet piled solution which would satisfy temporary and permanent conditions.

Contractor’s tender submission – Sheet pile wall to act as a temporary and permanent wall.

The issue

Proving a sheet pile will last for 120 years is a bit tricky. The River Thames is fresh (ish) water but the fact sewerage is being discharged nearby creates a risk of microbial induced corrosion which erodes steel and makes arguing the case that all will be fine (as in only 1 or 2mm rather than 5 or 6mm difficult). Unfortunately the UK national annex is more onerous than the main eurocode, I.e. assumes more corrosion occurs. 

Plan B

The contractor is keen to use a cofferdam which works in both temporary and permanent cases so have asked the Arup-Atkins joint venture to look at tubular piled options. Steel with concrete encased inside. This can be split into 2 cases. 1. Steel is non-sacrificial and last 120 years and 2. It does not.

Tubular pile option – outline sketch only

How does the cofferdam get built?

1. Modelling the construction sequence for either case is the same:
2. Install cofferdam
3. Install top prop
4. Dewater
5. Excavate to formation level
6. Install base slab (acts as a bottom prop)
7. Wait 2 years while shaft is excavated
8. Move to drained conditions.
9. Re-fill cofferdam/construct internal chambers to underside of prop level, back to undrained conditions.
10. Remove temporary top prop.
11. Finish cofferdam/structures, apply surcharge
12. Excavate and place scour protection.
13. Allow for some worst case tidal and tidal lag.
14. Long term case – move to drained conditions.
15. Case 1 – steel remains, Case 2 steel does not.

600mm tube option shown below.

Stage 5-8 in stages above – excavation to full depth with base slab at drained conditions

Final stage with deflection of wall shown in red (105mm) and bending moment in green (1MN)

Stiffness

Understanding shear, bending and deflection throughout the sequence was key. Various pile diameters were tried but this was simply an adjustment of stiffness in the Frew model. EI based on section size, steel and concrete Young’s Modulus. 0.7EI and 0.5EI were used for temporary and permanent concrete strengths. The difference for the 120 year case comes at the final stage where only the section properties of a concrete section above bed level can be taken into account, i.e. a reduction is stiffness. More onerous is the reduction in moment capacity the section has without a steel pile included.

The results

After trying various tubular pile sizes I got to a 600mm pile for both cases above as worst moments were in temporary case (max excavation rather than maximum fill height), so loss of steel section was not an issue as final stage was still below moment capacity.

Unfortunately I made a stupid error in the fill material (somehow I assigned a granular fill a high cohesion value) within the cofferdam which meant I had lower bending moments in the long term than I should have. That was identified before presenting to the contractor so I only lost face internally.

The amended results showed that a 600mm diameter pile works in the non-sacrificial case and a 900mm diameter pile works in the sacrificial case (due to reduced moment capacity). 

End state deflections were somewhat high (100mm) but all of the modelling was done at DA1 combination 2 which is not the case at SLS. Also some of the water profiling is a little unrealistic (high level inside the cofferdam, very low tide outside). There is, however, some assumed stress lockins. For example early stages see the wall bend inwards, later stages see it pushed outwards. Difficult to know if this will actually occur so this is a probable case and greater deflections might be expected.

Cladding

The front face of the river wall has stone faced cladding panels which need to be fixed to the piled wall. The connection detail raises questions on durability – casting in stainless steel brackets works but if you touch a stainless steel bracket against a mild steel pile you get greater corrosion so some thought is needed to avoid that. Additionally, final deflections up to 100mm although equate to a small angle mean the top of the cladding is off-set further in order to backfill between the cladding panels and wall. This means the cladding load acts at a greater distance causing more deflection. It also makes it harder to pour something in between the piles and the cladding panels. 

Design meeting

After a week of looking at all the options the proposed solution from Arup-Atkins (us) was the 600mm pile thick enough to corrode and be acceptable in the long term. I.e. a section size that was acceptable in the long term and then a corrosion allowance added. The contractor is now going to price this option. Although more expensive than a sheet pile option, it is cheaper than doing the vast sheet pile cofferdam and there are significant savings in temporary and permanent works (fewer internal props and ties needed because of the stiffer piles).

There was sensible input from the contractor about ease of installing 600mm diameter piles versus 900mm as well as lots of points about presenting a solution which meets the client’s durability concerns.

Outcome

We wait and see what Tideway say – they may reject all options and say build it as per the tender documents. If this is the case then I wonder what the point of early contractor involvement is.

Lessons

The whole issue and reason for having to look at alternative options is because of durability concerns raised by the client. The optimised contractor solution was/is a sensible plan and on a pain/gain contract it is worth trying to make savings.

The best thing I see having gone once around the process is developing multiple options to give the contractor the choice of what to present to the client. This enables a solution to be put forward on technical merit.

The plan to get the tubular pile costed before presenting to the client presents risk (for the contractor) has failed because of how busy the commercial team is right now which might turn out to be an issue. 

I have learnt a lot about the importance of how something is to be done without the need to go into detailed calculations – certainly the case at concept design stage.

Temporary works jetty solution to build the new site. All by river approach demonstrated with barge removing spoil.