studying up on seattle’s university link light ...coluccio.com/assets/uploads/files/university...

journalwww.tunnellingjournal.com

north american

EVERGREENLINE AWARDED

JUMP START ONAUSTIN’S JOLLYVILLE

AND OTHER NORTHAMERICAN NEWS

CABLE CRISIS?ARNOLD DIX LOOKS

AT UL’S WITHDRAWALOF ITS EMERGENCY

SYSTEMS CABLECERTIFICATION

CANADA’S BIGRIGS PAY OFFPLUS OTHER NEWS

AND CAREER MOVESFROM THE NA

TUNNEL INDUSTRY

TunnelingOct/Nov 2012

DETAILS ON PAGE 22SEE PAGE 5

UNIVERSITY CHALLENGESTUDYING UP ON SEATTLE’S UNIVERSITY

LINK LIGHT RAIL PROJECT

SEE PAGE 19

NATJ_1012_NAFNTCOVER_001.qxd:cover 21/10/12 20:44

TUNNELING IN SEATTLE is challenging. Andthat’s not just down to the region’s complexglacial geology. Client and contractor mustalso deal with the local press and publicperceptions of tunneling – vitally important ifyou need to secure funding for subsequentphases of work. For owner Sound Transit, theconstruction of the University Link (U-Link),the latest leg of its Link Light Rail network, is achance to demonstrate that it has learnt fromthe problems encountered on the previousCentral Link phase. And with the tunnelingcontract for the next light rail extension, theNorthgate Link, out for industry review, U-Linkcontractors will be hoping their achievementswill hold them in good stead.

As the U-Link tunneling contracts draw to aclose, NATJ spoke to some of the main players

to find out how lessonslearned on the Central Linkhave helped on UniversityLink – and what lessonscan be taken forward tofuture phases. Thoughthere has been a lot lessmedia publicity associatedwith the U-Link compared

to various other tunnel construction projectsin the region, the designers and contractorson U-Link have faced some formidablechallenges in the planning and execution ofthe project. For Joe Gildner, Sound Transit’sExecutive Project Director on the U-Link, thebiggest challenge was the unavoidableelement of the unknown. “On any tunneling

NORTH AMERICAN TUNNELING JOURNAL 11

Seattle’sUNIVERSITY CHALLENGE

U-LINK

The University Link extension to Seattle’s LinkLight Rail system has received less publicity thanthe previous Central Link section – but there havebeen plenty of challenges, quietly tackled.Kristina Smith reports

Top: Cross pas-sage breakouton U220. Left:JCM crews laythe invert forone of theU230 tunnels

NATJ_1012_011_018_Ulink.qxd:Feature 21/10/12 23:46 Page 11

U-LINK

12 NORTH AMERICAN TUNNELING JOURNAL

job, the challenge is usually how well thenature of the soils are going to play out andhow well the production is going to be for thecontractors,” he says. “On this job, althoughwe feel the design team did an excellent job inputting together a good GBR [GeotechnicalBaseline Report], we had three TBMs and 21cross passages to mine using SEM. And until

the machines show up, you cannot be sure.”Running for 3.15-miles (5km) between

Downtown Seattle and the University ofWashington to the North, the U-Link extensionis part of Sound Transit’s phased plan to createa light rail network in Seattle. Due to open in2016, University Link follows on from thealready-complete Central Link which runs fromSea-Tac airport to Downtown Seattle.

Next will come the 4.3-mile (6.9km)Northgate Link running from the University ofWashington up to Northgate Mall, 3.5-miles(5.6km) of which is below ground, withtenders due next year. The East Link, fromSeattle to Bellevue in Washington, includes asection of cut-and-cover tunnel in downtownBellvue. This is in its final design phase, withconstruction planned to start in 2015.

University Link runs completely undergroundin 21ft (6.4m) diameter twin tunnels, with thetunneling work split into two contracts, U220and U230 (see Figure 1). Let under design-bid-build contracts, the North Transit Partners (NTP)JV, consisting of Jacobs Associates, HNTB Corpand Earthtech, with 26 subconsultants,designed the tunnels and stations to 100%.

On U230 the Jay Dee/Coluccio/Michels JV(JCM) used a Hitachi Zosen EBPM to mine the3,610ft (1.1km) stretch between Capitol Hillstation and the Pine Street Stub Tunnel (PSST),

using the same machine for both tunnels.Traylor/Frontier-Kemper JV (TFK) used twoHerrenknecht EPBMs to construct the 2-mile(3km) long tunnels from Husky Stadium at theUniversity of Washington down to Capitol Hillfor contract U220.

The contracts to build the station have beenlet separately, in contrast with the combined

tunnel and station contract employed on theBeacon Hill section of the Central Line. OnBeacon Hill, substantial completion was 11months late on top of 70 days of agreed timeextensions.

Both U-Link tunneling contracts involvedretaining works for the station boxes, withJCM excavating the whole of Capitol Hill, andTFK excavating launch shafts at the Southern

end of the University of Washington Station.“For us the focus was coming off Beacon

Hill,” says Gildner. “We wanted to have tunnelcontractors who could stay focussed on theheavy civil underground construction activitiesand give ourselves the opportunity to look foranother contractor to focus on the stationactivities.” This approach allowed SoundTransit to procure the tunnelling contracts atcompetitive prices – it was reported at the timethat the U220 contract had come in 22%below the Engineer’s Estimate.

It also meant that Sound Transit could use aGCCM (General Contracting/ConstructionManagement) approach to the procurement ofthe station finishes. “GCCM gives us the abilityto hire a contractor through a qualificationprocess,” says Gildner. ”And we can then gethim in here to help with constructability andvalue engineering of the station layouts.” Onthe Northgate Link, this approach will be takenfurther as the station finishes will be only 60%designed, allowing for innovation and savings.

Of course, more contracts mean moreinterfaces for the owner and constructionmanager START JV (CH2M Hill and JacobsEngineering) to manage and there have beenissues. For example, TFK was due to takepossession of the Northern end of JCM’sstation box and some adjacent land in order todismantle its two TBMs, which broke throughinto the box at the end of their runs. However,JCM were still tunnelling, using that area fortheir muck bin and having to hand it overwould have caused them serious logisticalheadaches. A solution was found that suitedeveryone, with TFK’s two machines waitingpatiently for four months to be dismantled.

“TFK, JCM and Sound Transit came togetherand worked through the necessary interfaceissues,” says Gildner. “I am real proud of theway all three parties stepped up.”

While the University Link tunneling workwas split into two contracts, the NorthgateExtension will be let in a single contract,estimated between $500 and $540 million

Portage

University ofWashington

Station

Capitol HillStation

PSST

Bay

LakeUnion

5

5

5

520

Tunnel route

Station

Light railto Airport U230 Cont

rac

t

U22

0 C

ontr

act

Montlake CutCrossing

North Corridor HCT Lynnwood

North Link

University Link

Northgate

East Link

Overlake

DowntownBellevueDowntown

Seattle

Downtown Seattle to Sea-Tac Airport

University ofWashington

Sea-Tac Airport

South Link

Redondo/Star Lake

Final Design TBDConstruction TBDOpen for service 2023

Final Design 2006-2008Construction 2008-2015Open for service 2016


Final Design TBDConstruction TBDOpen for service 2023

Open for service July 2009


Roosevelt

Brooklyn

Capitol Hill

StadiumSODOBeacon Hill

Mount BakerColumbia CityOthelloRainier BeachTukwila Int. Blvd.

Tunnel route

WestlakeUniversity Street

Pioneer Square

Int. District/Chinatown

N

Figure 1: Schematic of Seattle’s future Link Light Rail system and detail of U-Link



incorporating the running tunnels, 23 crosspassages, two underground cut-and-coverstations and a portal at the north end.

“We have even smaller job sites on theNorthgate Link extension than we’ve had towork with on University Link,” says SoundTransit spokesman, Bruce Gray. “Based on anextensive value engineering process, we arrivedat the single tunneling contract approach toavoid interface issues and hopefully save onoverall construction costs with more efficientequipment and overhead use.”

No chance of voidsOne of the biggest lessons that Sound Transitlearned on the Central Link project related toover-excavation by the TBM as it was relauchedfrom the east end of the station. Eighteenmonths after tunnelling on the Beacon Hillsection was complete, a local homeowner wassurprised to find a 12in (300mm) hole in herflowerbed adjacent to the front porch, which –on closer inspection – proved to be the neck ofa 23ft (7m) deep, by 25ft (7.6m) diameter,dome-shaped cavity within 3ft (1m) of thecenterline of the 110ft (35m) deep NB tunnel.Subsequent investigations located furthervoids, which were filled with controlled densityfill (see NATJ, June/July ‘11, p14).

One of the measures employed by SoundTransit to guard against over-excavation onBeacon Hill was the use of muck scales on theconveyors. However, the contractor hadexperienced problems with them prior to therelaunch of the TBM from the station, so spikesin muck weight were written off as amalfunction of the equipment.

Speaking to the contractors from both U-Link contracts, it seems that problems with themuck scales still exist. Both found them to beless than reliable. “Scales can be accurate if thematerial going over the scales is consistent, forexample in gravel quarries,” says Glen Frank,Project Manager for JCM. “When the materialis consistent, you can calibrate the scale so thatyou they are measuring very accurately. When

ground is changing, one minutealmost liquid, the next big chunkscoming through, you will not getaccurate results.”

TFK discovered that the scales couldnot weigh anything over 90lbs (40kg).So when big lumps of hard clay weretravelling over the conveyor, theweights of muck recorded werestrangely low. “We were scepticalthat they were going to work,” saysTFK Project Engineer Matt Burdick.“But at the end of the day, althoughthey weren’t very useful in terms ofproviding discreet data points, theydid provide a windsock. Over a periodof time, over 150 rings, you develop atrend.”

For both contractors settlementcontrol relied more on the real time monitoringof pressures, watching the EPB sensorsbetween drives to make sure sufficient pressurewas maintained – and grouting through thetail shield immediately.

Frank explains: “When we apply pressurethrough the machine, we change the porepressure up into formation. It depends on thepermeability of the formation as to how muchtime it takes for that pore pressure to bleed offand come back to constant pressure. If, whenyou stop mining, the pressure in the plenumdrops back down, you are applying enoughpressure. If it does not, then you are not.”

For Gildner, it was important that both clientand contractor were watching all theimportant areas: Muck weights, pressures andsamples if required. “Both parties monitoredthe face pressures and the grout pressurescontinually in real time,” says Gildner. “Therewere metrics we required shift by shift so thatboth parties were keeping tabs on these.”

Any anomalies required action. “If we had

two or three rings with higher muck scaleweights and lower grout takes, we had to goback and proof grout,” says Burdick. “That didnot happen very often. “If we noticed we hada high muck scale weight we would sit thereand grout to pressure, maybe push the grout alittle harder to fill up any potential gaps. As wehad up to 300ft (90m) of cover, there was nopotential problem with heave.”

Another strategy employed by Sound Transitto ensure the smooth running of the U-Linkcontracts was to carry out a series of advancedworks including securing some permits,demolition works at the station sites anddiverting utilities. One of the most technicallydemanding advanced works contracts sawCondon-Johnson & Associates clearing apathway under the I-5 highway, which liesabove the path of both JCM’s drives. Condon-Johnson had to cut four holes through theretaining walls’ piles, install tie backs to helpstabilize the walls, backfill with controlleddensity fill and install monitoring on the walls.

Workers celebrate the completion ofthe U230 drives between Capitol Hilland Pine Street in downtown Seattle



Contract U230: Capitol Hill to the PineStreet Stub TunnelWith the technical challenges on ContractU230 now behind him, Frank deserves to feelpleased. “Everybody agrees that this projectwas a success,” he says. “We are going tomake a profit. The owner is on budget and onschedule. And there were no impacts on thirdparties that were non-negligible.”

There were plenty of issues to think about:the runs were downhill, so JCM had to ensureit had sufficient pumps to clear water fromthe heading; almost the entire length of bothtunnels were on curves, at some points in twodirections; the ground changed “radically andoften”; and passing under 32 buildings withlimited cover meant control of settlement wasparamount. The owner-controlled insurancemeant that JCM had to pay $100,000 ofdeductibles on any damage. “We were veryconservative in making sure we werecontrolling the ground. We did some extrasettlement monitoring and measurement,”says Frank. JCM’s Hitachi Zosen machine hadan extended overcut, partially to limit wear onthe cutter tools, but also to accommodate thedual component grouting system with itsinjection ports outside the machine’s shield.JCM had to do a proving section todemonstrate that the overcut would not leadto settlement.

But of all the challenges, the biggest forboth designers and constructors on contractU230 was the section where the tunnels rununder the I-5 highway, at its closest just 14ft(4m) below the expressway. “The I-5 is animportant transport route for all the goodsand services that go up and down the Westcoast,” says Peter Raleigh of JacobsAssociates. “If you are going to do any

damage you would really be in trouble.”Theanswer to this challenge came in thespecification for the TBM, says Raleigh. “Youneed a reasonable amount of control of themachine, it’s got to be properly equipped withproper instrumentation.”

“We had to prescribe that once thetunneling started, they had to stay at itcontinuously,” adds Gildner. In parallel withthe TBM considerations came the requirementfor instrumentation and measurement at thesurface. There were reflective targetsmonitored by theodolites from a distance andhorizontal inclinometers over each drive,although the latter did not always perform asexpected, says Raleigh.

For the tunnelers, the stretch under the I-5was a tough one too. They needed to addplenty of conditioner to the denseimpermeable clay in order to process the

muck. “Our screw conveyor and muckhandling system did not react very well to claythat remoulded,” says Frank. However, underthe highway, the contractor had to limit theamount of foam going in – since there wereborings up to the surface through which thefoam could have escaped.“We had twocompeting goals,” says Frank. “The operatorswanted to make production – we all want tomake production – but the engineering staffhad another goal of making sure we did notcause any damage on the surface.”

They had already seen the results of foamsqueezing up through a borehole on an earliersection. A slimy mixture of foam and silt hadshot 40ft (12m) into the air, dirtying a tree andtwo parked cars. Fortunately, there were nopeople in the vicinity at the time.

Slimy mud on the road would have been ahazard for drivers, but was something JCMwas able to avoid. Some foam did make it tothe surface, but thankfully emerged in agrassy embankment and ran down only ontothe hard shoulder, which the contractor wasable to clean during the usual night timeclosure of the highway.

“The contractor did a very good job onthose two runs,” says Gildner. “And I wasreally pleased with the way the contractor,owner and DoT came together.”

The tunnels’ curves, the tightest being a550ft (168m) radius, complicated the ringbuild. The contractor had to add a thirdsegment design, on top of the up rings anddown rings, which had twice the taper of theothers. “We had to spend more time trainingthe crews, to make sure we could build therings so they butted up correctly to eachother,” says Frank.

Both contracts used the same segments,cast by a Traylor Bros and Technopref jointventure using existing moulds that came from

U-LINK

Existing vault and utility line

Sewerline

Southbound tunnel

Northbound tunnel

Existingpiles

Existingtiebacks

ExistingPSSTheadwall

Pine St

Angled GroundTreatment Zone

Figure 2: Jet grouting at the PSST

Far left: Finalbreakthrough cele-brations as JCM’sHitachi ZosenEPBM completesher final drive

Left: The end wallformwork isinstalled on U230

Below: Completionof cross passage 17on U220


U-LINK


the LA Metro project.One of the most complex areas of the U230

contract was where the two tunnels joinedthe existing Pine Street Stub Tunnel (PSST), notleast because the tunnel was in usethroughout the works. NTP made extensivechanges to the final design, which SoundTransit’s preliminary designer ParsonsBrinckerhoff had proposed, and the fact thatthe contractor followed the new design withlittle adjustment is testament to its success.

The original design included two receptionshafts, with the last stretches of both tunnels,around 100ft (30m) long, up to the retainingwall of the stub tunnel constructed usingSEM. The difficulty in this area was thepresence of soldier piles and tie backs –temporary works from the Pine Street StubTunnel construction – together with otherunderground obstacles. Another challengewas that the tunnels had to break through theretaining wall at angles rather than straighton. This was partly due to an earlier change,which moved the station to Capitol Hill fromFirst Hill – abandoned because it would havebeen a deep station in difficult ground, similarto Beacon Hill.

NTP’s reworked design saw extensiveground treatment of the area using jetgrouting so that the TBMs could just tunnelright through. There was one reception shafton the Northbound tunnel with a hand dugadit connecting to the Southbound tunnel, sothat some parts of the machine could beretrieved at the end of the secondSouthbound run, leaving the TBM’s shield inthe ground. NTP also came up with the ideaof a false wall inside the stub tunnel,separating the ongoing operation in thetunnel from the activity and dust of theconstruction operation. The new solution hadsafety and cost on its side.

“We pretty much built it per plan,” saysFrank. “We made the adit a little bit bigger,which allowed us to save more of the TBM.We had to change the shaft excavation andremove a couple of piles in front of the adit sothat we could get bigger pieces of themachine out.”

Specialist contractor Malcolm Drillingcarried out the extensive groutingprogramme, using large diameter overlappingjet grouted piles, to stabilize the groundbehind the stub tunnel headwall (see NATJ,Apr/May ‘12, p19) and Figure 2).

There had been concerns that JCM wouldhave to put men into the cutterhead to cutthe ties, the worry being that the ties couldwrap themselves round the cutterhead,preventing the machine from advancing. TheHitachi Zosen munched them up nicelyhowever, says Frank, producing 2ft longindividual strands, and with just a fewwrapped round its cutterhead as it brokethrough into the stub tunnel.

“The TBM was hitting the ties end on,”points out Raleigh. “If the TBM had been

running perpendicular to the ties, it mighthave been a different story.”

Having tackled the design issues around theconnection at the Pine Street Stub Tunnel, NTPhas made sure that they aren’t re-encounteredwhere the Northgate Link will join to theUniversity Link. So at the University ofWashington there will be a hidden chamberto receive the TBMs without disturbing thestation’s operation; fiber glass reinforcementin the headwall where the TBMs should breakthrough; and the headwall will be angledperpendicular to the anticipated TBM drives.

While at the Pine Street end of the jobthings ran smoothly, in other locations thingsweren’t exactly as expected. It was water – orrather its position – which caused the biggestsurprises. First, the contractor discovered as itexcavated the Capitol Hill station box thatthere was much more water than expected.“Rather than the bottom of the permeableground being 15ft to 20ft below the bottomof the excavation, it was above the bottom ofthe excavation,” says Frank. The quickestsolution was to put more dewatering wells in,but the excavation ended up taking sevenmonths rather than the intended five.Thankfully, the problem was limited to oneend of the box however, which meant JCMcould pour concrete at the other end and startassembling the TBM.

A second, ongoing, differing site conditionsclaim relates to the construction of two of thefive SEM cross passages, where the contractorclaims ground conditions were more onerousthan expected. “Two were designed to bedewatered but we ended up with three thatneeded to be dewatered,” says Frank. “Thedifference was in where the aquicludes were

in relation to the aquifers.”Subconsultant Dr. Sauer Group designed

the 21 cross passages for both contracts. “Thenotable thing about the design was that thelocation of the cross passages was chosen inorder to minimize the adverse impact of theglacial soils, placing them in the morecompetent soils, rather than in the sand and

gravels,” says Dr Sauer’s MD JuergenLaubbichler. So the final position of the crosspassages was not fixed until the groundconditions had been inspected at that point,stopping the TBM to do so. The contractrequired these inspection stops which TFKperformed with a camera mounted inside theTBM bulkhead. If the camera inspectiondetermined conditions were safe, theexcavation chamber hatch was opened and asoil sample was taken.

Older glacial deposits (Pleistocene): QPGTM-Pre-Fraser subglacial meltout till deposits

QPGF-Pre-Fraser coarse-grained glaciofluvial deposits

QPGL-Pre-Fraser fine-grained glaciolacustrine deposits

QPGT-Pre-Fraser glacial till

QPGD-Pre-Fraser glaciolacustrine diamict

Vashon glacial deposits: QVT-Vashon subglacial till

Non-glacial deposits (Holocene):

AF-Fill

QAL-Alluvium

NB 1196+00 NB 1197+00 NB 1198+00 NB 1199+00 NB 1200+00

80

60

40

20

0

-20

-40

-60

-80

80

60

40

20

0

-20

-40

-60

-80

Existing ground level

Water level

Lake Washington Ship Canal

Tunnel

Figure 3: Alignment of the U220 tunnels under the Montlake Cut

Preparation for invert casting on one of theU220 cross passages


U-LINK


The design foresaw some cross passages inCategory II ground. “There was an excavationclass that prescribed probing and vacuumdewatering in combination with pre-supportand pocket excavation to ensure the stabilityof the face and of the tunnel excavationitself,” says Laubbichler. “After an initiallearning curve, and with guidance from DrSauer SEM field personnel, this was executedvery well by both contractors.”

Despite the technical issues on U230, thesehave not offered the biggest challenges toFrank and his team. The biggest lessonlearned was about working with SoundTransit. Next time they would allow for moreresources to deal with the high flow ofinformation required from contractor to client.

“The challenge that we felt we had waskeeping our scope to just building thetunnels,” says Frank. “We felt like we wereasked to provide an awful lot of consulting tothe owner. Because it is so urban, there aremultiple third parties and the owner, being atransit authority, wanted to communicatewith all of them. Now we pretty much knowhow they are going to administer that nextcontract. We have a good feel for how muchsupport we need to supply them in order forthem to be able to take care of their interestsand how much staff time and effort we needto be able to supply.”

Contract U220: University of Washingtonto Capitol HillTime was the biggest enemy for TFK oncontract U220. Their site at the University ofWashington end was basically borrowing landfrom the Husky Stadium’s car park, so thelease they had was as tight as it could be.“Having the construction site on the HuskyStadium doorstep was a real sacrifice for theUniversity,” says Raleigh. “For the contractor itmeant that they had to keep their site veryclean, neat and tidy.”

The impact went far beyond goodhousekeeping, however. The really interestingresult of the schedule was that TFK had tostart mining cross-passages whilst it was stillconstructing the running tunnels. The solutiondevised has provided good lessons, saysBurdick – on what works well, and whatworks less well. The first innovation saw TFKemploy Halcrow to design an alternativepropping system to the squirrel cages, whichwould traditionally be used. This involvedtemporary shotcrete on the tunnel lining,around the area where the segments for thecross passage were to be broken out.

This has worked well, but what Burdick calls“the 600lb gorilla in the room,” is how hard itwill be to break this temporary shotcrete off.When NATJ spoke to Burdick TFK waspreparing to start the removal operation thefollowing week, armed with an array ofmachines and tools, not knowing yet whatwould be most effective. “I have multipleconcerns,” he says, among them safety, ease

of operation, quality of final result andprogram. Assuming the removal operation issuccessful, this form of propping could beuseful on future jobs, says Burdick. “Whetheryou used it would be a function of how manycross passages you have to do in theschedule,” he says. “We needed to have a lotof propping installed for a long period of time.For us it was cheaper to use shotcretepropping rather than steel because of theexpense of fabricated steel. With shotcreteyou are able to install it in around three and ahalf shifts. Steel takes twice that long, it’s veryexpensive and pretty dangerous.”

The second innovation for the crosspassage construction was collapsible workdecks which folded down onto a scissor car.These served as working platforms for miningof the passages, allowingthe locomotive bringingsegments to passunderneath. A miniconveyor carried muck offthe platforms to join themain conveyor.

“The biggest problemwas that we were workingon up to five crosspassages at some stage atany given time, behind themachine that was 1.5-miles ahead and stillmaking 70ft a day,” saysBurdick. “Work thatrequired blocking the rail,such as work deckinstallation, had to besequenced around thetrains feeding the TBM orperformed over theweekend during non-mining shifts. It was a huge co-ordinationeffort.”

There are things Burdick would dodifferently another time. “The actualexcavation is difficult because until themachine stopped, the vast majority was doneby hand which is 1890’s technology,” he says.“With a little bit of change to the design wewould have been able to do moremechanically. It could be done.”

Gildner is appreciative of TFK’s methods forthe cross passages.”It was very innovative ontheir part,” he says. “You have really got totake your hat off to them to be able to dothose two major activities at the same time.”

When Raleigh was asked whether hewould advocate constructing cross passagesat the same time as running tunnels again,this was his response: “Contractors can makemistakes if they are doing something that isunconventional or uncomfortable. But youcannot always have harmony betweenschedule restraints and the use of standardmethods.”

For Burdick, however, the biggest technicalchallenge was not the cross passage

construction, but preparing for somethingthat never happened. The highest expectedpressure along the route was 6.2 bars, whichmeant that TFK’s machine and processes hadto be able to cope with interventions toinspect the cutterhead or change tools.

The first hurdle was gaining the necessarycompressed air working variations from thestate. Laws relating to working at highpressure have not been updated for years, sospecial permissions are required to workabove 3.5 bars. The Herrenknecht machinewas also set up with a double manlock so thatcrews could work in rotation if necessary. Andthe contractor had a shuttle lock to ferryinjured compressed air workers to the onsitemedical lock for treatment or deliver diversfrom the surface under pressure, should

saturation diving be needed for longerduration interventions. Happily, nointerventions at high pressure were required.The massive boulders, which the contractorsthought might have to be broken up usingmen and tools, did not materialize.

In fact, the cutterheads emerged at theother end with some paint still intact. “I don’tthink I’ve ever seen cutter wear better thanthat,” says Burdick. “Essentially it was likecutting through grease – the clay was alubricant.” The contractor will now sell thetools onto the Traylor/Skanska/Jay Dee JV onBlue Plains, Washington.

On each of the drives, TFK did make toolchanges, both at the same point – before themachines entered a sand-filled valley, thoughthey were not needed, says Burdick. Only thenose bar and the raker on the nose barshowed significant wear. At this point it wasalso important to check the cutter disc seals,says Raleigh, to ensure the bearings that allowthe discs to rotate are still protected.

The high head of water, though it didn’tcause problems for interventions, did causesome issues with the two-component

One of TFK’s two Herrenknecht EPBMs


U-LINK


grouting. The Herrenknecht injection systemvaries from that used on the Hitachi Zosen inthat its injection ports are inside the shield.

“This is the highest pressure that we havedone bi-component grouting with,” saysBurdick. “We found that at those pressuresthe grout is forced back into the tail skinbrushes. You need good tail skin brushes andgood tail skin grease. We learned a few thingsthere.” Having had to replace a back row ofbrushes which got grouted up, and tryingevery tail skin grease known to man, Burdickconcludes that it is really important to usehigh quality tail skin grease and apply itreligiously at the right dosage. He also advisesthat foam tape applied along the radial jointsof the segments along the ring end makes a

significant difference in limiting grout leakage.And he advocates a clean ring build area toavoid getting dirt on the gaskets which impacton the seal achieved.

“That’s one of the things we need to workon, figuring out a better way of sealing themachines,” says Raleigh. “There aredevelopments underway in that area to comeup with a differing type of sealing betweenthe tail shield and the rings.”

As on contract 230, TFK are claimingdiffering ground conditions at certain cross-passage positions. Of the 16 they had tobuild, four were expected to be in Category IIground and require some dewatering. But TFKclaims the ground was worse than expectedat cross passage 17, requiring dewateringfrom the surface before completion, anoperation which was not intended in theoriginal design. Just one cross passage, 9, wasrelocated to better ground after the inspectionvia the TBM.

One area which looked potentially tricky on

paper, but which passed without incident,was the section at the beginning of each drivewhere the TBMs set off and headed downhillto burrow under the Montlake Cut, ashipping canal connecting Lakes Union andWashington. NTP had adjusted the alignmentfrom the preliminary design, reducing thedistance between the top of the tunnel andthe very lowest point of the Montlake Cutfrom 23ft (7m) to 13ft (4m), incorporating 8ft(2.5m) of the hard clay and 5ft (1.5m) ofsediment at the bottom of the cut. The reasonfor moving the tunnel up was to optimize thegrade and also reduce the depth of theUniversity of Washington station, reducingcosts. “If that had gone bad, it could havebeen very bad, if we chimneyed all the way to

the surface or got a huge blow in,” saysBurdick. With a team of spotters poised onthe surface, the only evidence that anyonesaw was a few bubbles rising to the surface,possibly conditioner or possibly gas fromrotting vegetable matter at the bottom of thewater course.

Apart from the interface issue where TFKhad to wait to get into the Capitol Hill stationboxes, there have been few contractual issues,says Burdick. “It is a $300 million job and wehave on the order of 2% or 3% change incontract the value, which is pretty good.”

Looking ahead: Northgate LinkThere are many practices which Sound Transitis taking through from U-Link to NorthgateLink. Though the owner is looking to newways of procurement, for example usingGCCM on the station fit out contracts, itprefers to stick to Design-Bid-Build fortunneling. “Together with the way we useGBRs, Dispute Resolution Boards and Escrow

documents, we feel very comfortable withDesign-Bid-Build in tunnels,” says Gildner.“This is mainly from the standpoint of ourallocation of risk when it comes to heavyunderground civil work.”

Gildner believes that it is really important toinvest sufficiently in producing a good GBR,drawing on the empirical data from allSeattle’s tunneling projects. “As we frame thenature of the hydrogeological conditions, wehave to make sure that we are providing anadequate baseline and make sure the biddersfeel comfortable with the way we havecharacterized the soils.”

Sound Transit’s method of selecting thethree-man Dispute Resolution Board jointly,rather than each party electing someone,works well for them says Gildner. On U-Link,some differing site conditions claims had togo before the DRB, although others wereaccepted without it.

Another practice which Sound Transit willtake forward is specifying the primary supportsystem for SEM tunnels. On U-Link, therewere two categories of primary supportspecified as a result of the hydrogeologicalconditions. To this the specification adds anadditional ‘toolbox’ of supplementary items.

“The SEM experts down at the headingmight want to call for supplemental items,”says Gildner. “We list all these out and pre-price them at unit prices. We think this is fairand reasonable.”

There is one area where the specificationmight change. And that relates to the TBMs.The specification will call for a machine thatcan maintain active face pressure, as on U-Link. but Sound Transit has not decided yetwhether it will insist on new machines as ithas in the past. “We have been specifyingnew machines, based on risk assessment,”says Gildner. “We recognizes the importanceof the hydraulics and these soils to tend to beabrasive. However, the agency has yet to takeanother look at specifying new versus used.”

As for U-Link, there is still a long way to gobefore its opening in four years’ time. Gildnerestimates that the project is around 60%complete.

With tunneling work behind them, the riskregister must be considerably shorter than itwas. Both the crossings beneath the I-5Highway and the Montlake Cut were“earmarked as high risk items” says Gildner.

“I think we more or less got it right,” saysRaleigh. “Or I should say that the ground co-operated. Because what you have toremember is ‘the ground is the boss’. The testcomes when you don’t have good ground towork with. You need to stay as flexible as youcan be with the design and give thecontractors as much flexibility as you canwithout losing sight of the goal. Preparing thedesign is a balancing act between being veryprescriptive and being very performance-oriented.”

Shotcrete lining works by TFK on the U220 cross passages


studying up on seattle’s university link light ...coluccio.com/assets/uploads/files/university...

Documents