Construction Vibration
David Trevor-Jones
 |
|
| The church of Ethelburga, Bishopsgate, City of London with 100 Bishopsgate demolition site immediateley adjacent (All photos: David Trevor-Jones) |
It might seem to be rather obvious that an ancient structure is vulnerable to vibration. But church towers have been moving in response to bell-ringing since bells were invented. So why do we need to worry about vibration now?
The feature of the present era that has brought the question to a head, especially in the City of London, is adjacent redevelopment. The intensity of redevelopment has grown with the forest of cranes that has stood over the City since the early 2000s. Demolition of 1970s and 1980s buildings very often requires breaking massive concrete elements and thick reinforced concrete ground slabs. Huge amounts of energy are released into the ground, usually percussively, and sometimes insufficient thought is given to how far it might travel and what effect it might have.
Vibration, like sound, can be deflected and reflected. Man-made and geological discontinuities in the ground influence the path of the threedimensional sub-surface waves while a surface wave (called the Rayleigh wave) behaves more like the ripples on a pond. About 50 metres below the surface in the City the top of the London Clay beds represents a reflective surface to some ground wave modes. The result, spectacularly demonstrated during large scale demolition on the east side of Bishopsgate in the City at times over the past ten years, can be very long distance propagation. Vibration has been observed and measured at distances of the order of 100 metres from its source. It is not necessarily the development site next door that is generating the vibration detected in a building, and short-distance propagation might be weaker than long-distance.
Modern office blocks tend to require very deep piling and very deep, multi-level basement excavation. While resultant ground movement is probably a greater long-term concern for neighbours, vibration can be a hazard in the short term for surrounding buildings. And whereas the response of relatively simple, stiff, engineered structures to ground vibration is reasonably predictable, that of ancient structures is not.
ST HELEN'S, BISHOPSGATE
It was the demolition of the massively over-engineered, reinforced concrete Crosby Square building from the 1970s to make way for The Pinnacle skyscraper (since re-cast as 22 Bishopsgate and presently under construction) that brought these issues into the author’s domain. The demolition started to generate unexpectedly high magnitudes of ground vibration and alerted the neighbourhood to the unusual local ground conditions that seemed to propagate it. The medieval church of St Helen’s Bishopsgate was affected and when redevelopment of the immediately adjacent southern block of St Helen’s Place received planning permission as well, a fundamental review of the threat to the ancient structure of the church (albeit as substantially restored by Quinlan Terry after the 1993 Bishopsgate bomb atrocity) became pressing.
A vibration monitoring and management regime was included in the Neighbourly Matters Deed negotiated by the representatives of the church and the developer. The question was, how to specify meaningful, appropriate vibration level thresholds to trigger an escalating management response? The parties agreed to fund a limited literature search and consultation exercise to be conducted jointly by their respective expert advisors, the present author acting in that for the church.
 |
||
| St Helen's Bishopsgate (just right of centre) with St Helen's Place immediately to its left (before its demolition) and the demolition site of 100 Bishopsgate further left. |
Standards addressing the thresholds of exposure to vibration that could indicate a risk of building damage have been developed over more than 80 years, starting with work in Germany before the Second World War. Part 3 of the current German standard, DIN4150, is an important authority. So is the corresponding British Standard, BS7385-2, but neither addresses potentially fragile, ancient and non-standard structures other than rather unhelpfully broadly.
The outcome of a wide-ranging consultation with UK-based building conservation and special interest groups was that surprisingly little information was available. It seems that development including demolition has taken place close to ancient buildings without much attention to their vulnerability to ground vibration, not least in London. The sites yielding the greatest case study information were in Southampton, where the West Quay shopping centre development had impinged on the medieval city wall, and at Christ Church Greyfriars where the construction of an office building had impinged on both the church (maintained as a controlled ruin since destruction in the blitz) and part of London’s Roman wall.
Eventually the synthesis of standards, documented evidence (scant though it was) and anecdotal experience led to the relevant parties agreeing amber and red alert threshold levels along with an absolute limit value for vibration monitored at points on the medieval rubble north wall of St Helen’s.
The term ‘limit’ is widely used in noise and vibration monitoring work, but it is only meaningful if breaching it has a consequence. In the case of managing vibration to avoid building damage, the limit value must represent the onset of a real risk and therefore has a real significance. At St Helens, the amber and red threshold levels were established to help manage the processes generating the vibration, so that the absolute limit would never be exceeded. Amber alerts would be made available only to the construction site team to advise them that significant but sub-actionable vibration was occurring, escalating to red alerts, usually shared among all interested parties, to warn that vibration was approaching a significant risk level. A protocol was established to set out a hierarchy of actions to be taken in response, from noting and recording to suspending work to review methods.
The alert and limit values are specified as velocity values. Vibration may be measured as a displacement, velocity, acceleration or even jerk (acceleration of acceleration) and the proprietary analytical software packages now commonly available with monitoring instruments allow transformations between them to be made rapidly and easily. The advantage in referring to velocity is that the values are meaningful in both building damage risk and human perception applications. Geophones, which are velocity transducers, are often used for this purpose.
The alerts are sent out from vibration monitoring devices via mobile phone or wi-fi networks. The alert levels are programmed in and a message is automatically sent to a list of registered recipients if they are exceeded. As a rough order of magnitude indication, amber alert thresholds for churches are usually in the 2–5mm/s range and red in the 3–8mm/s range, with the absolute limit in the 5–15mm/s range. The choice of actual threshold depends upon the specific vulnerability of the subject building.
The positioning of monitors is guided by an International Standard (BS ISO 4866) but in the cases of historic buildings generally and of churches in particular, locations must be chosen to reflect the specific sensitivities of the subject. Monitoring on memorials fixed to the church walls can be helpful both because they can be seen in some degree as the canary in the coalmine in respect of the building’s response to vibration, and also because historic memorials are valuable works of art and sensitive to vibration in their own right. Although unlikely, the nightmare outcome could be a precious work of art ‘walking’ off its wrought iron fixings under vibration excitation.
Conventional masonry and especially framed buildings tend to sway in response to ground vibration, both because energy in the ground naturally rocks them about a rotational axis in or below their footings or foundations and because of the relief of the inertial mass imposed by higher storeys with height up the structure. The monitoring at St Helen’s was designed around that proposition with monitors placed high up and low down, and distributed along the length of the medieval north wall, including on one of the important memorials fixed to it. The results, which included documented incidents of serious impact and loading, showed that the random stone facing with a well-consolidated rubble core, all bonded with a lime mortar of some kind, reacted dispersively – that is to say, the vibration energy dispersed through the slightly elastic lime mortar, absorbed along the way by the frictional and dynamic losses in the wall much as water wave energy is dissipated by a boulder beach defence. The structure itself was initially investigated and a watch was then maintained by a specialist team of conservation professionals during demolition work. The wall did not sway as a modern one would and reacted to at least one major accidental impact just locally, the impact energy absorbed by the complex threedimensional web of mortar in the irregular joints between facing stones and bonding of the random rubble core.
 |
|
| Triaxial geophone mounted on a bracket high up on the north wall, St Helen's Bishopsgate, City of London | |
 |
|
| Triaxial accelerometre mounted on top of the Throckmorton memorial, at Katherine Cree, City of London. |
TIMBER STRUCTURES
The St Helen’s experience can be extended to timber structures. At St Mary Abchurch, adjacent to and partly over London Underground’s works to extend Bank Station, the principal aim of vibration monitoring was to protect Wren’s frescoed saucer dome ceiling and the timber roof structure supporting it. It was speculated that the sheer complexity of the timber structure would be key. Timber moves naturally under varying humidity, temperature and wind loading so to some degree the roof and the dome suspended from it have probably rocked and swayed significantly over the centuries (it even survived the Blitz, albeit with considerable damage to the ceiling). The key differences between natural perturbation and the forces that can be imposed by demolition and construction are that the latter can include sudden repeated shocks and relatively long-term repeated excitations.
A veritable forest of timbers in the roof of St Mary’s with very many complex joints provides a dispersive system similar in concept and probably in behaviour to the three-dimensional net of lime mortar bonding all ofthe elements in St Helens’ north wall. A vibration monitoring network was set up with a root monitor on the top of the masonry east wall that would provide measurements both of any motion available to transfer into the timber structure and at the same time, of any sway mode in the wall itself. Other monitors tracked vibration through the roof timbers to the rim of the dome. To date, the assumption that any imposed vibration measurable on top of the wall would disperse through the timber structure before reaching the dome appears to have been correct.
WORKS OF ART
Despite the dispersive effect of traditional timber framed structures, the risk remained that the frescoed dome at St Mary Abchurch could be affected by residual energy to some degree. How might works of art such as this respond to and be protected from vibration? Major art museums around the world have had to tackle the dilemma of how much of their collections can remain on display safely when major construction works, often to extend the museum’s own buildings, go on around them. This has been a matter of concern at the National Gallery, British Museum and Victoria & Albert Museums in London. The same issues have also been addressed at museums in the USA, and an interesting study has been conducted in Florence of the response of Michelangelo’s ‘David’ to vibration from pedestrian footfall in the gallery around it. The issues highlighted in research reports include attritional damage to painted gessoed timber (ancient Egyptian coffins seem to be singly the most vulnerable museum artefacts) and failure of old repairs. ‘Walking’, the vibration-assisted movement of artefacts off shelves is also highlighted as a real and potentially catastrophic risk.
The relevance of the reports and papers on these museum collection studies to the British parish church is that our historic churches are themselves museums of works of art, from frescoes and murals to framed paintings on canvass or panels, commissioned and donated works of art and craft, statuary and memorials, and many others. Potentially these are all vulnerable to the effects of vibration.
THE IMPORTANCE OF FORWARD PLANNING
The most effective monitoring process is one that starts early. There is no British Standard historic church and there can be no standard handbook for designing a monitoring network for a particular building in its own particular circumstances. Neither are there standard alert thresholds or limits unless the structure is relatively simple, well-understood and in good condition. Every historic church will be different and must be evaluated and treated on merits. The input of a conservation architect advising on its likely structure and condition is invaluable. Every morsel of information from those familiar with the behaviour and quirks of the building, from the fabric warden to the quinquennial surveyor, is useful in determining its specific vulnerability and therefore the right regime of alert levels to select to protect it.
No discussion of this kind is useful if the church concerned does not have the funds to pay for advice. That is where well-negotiated Neighbourly Matters Deeds are likely to be vital. The cost of setting up the kind of monitoring exercise outlined above will be in the low thousands, with an ongoing recurring weekly cost in the several hundreds of pounds. That cost would be an unsupportable burden on the average parish church, but it is likely to be minor in the context of a major construction project. On the other hand, the potential cost to a developer of causing irreparable damage to an historic building or work of art in consequent claims and reputational damage could be vast.
The reality is that the alerts system is expressly intended not to impede work on site but to help with its management. A site manager with confidence in the system can adopt and respond to it in such a way as not to incur stoppages and delays. A well-conceived system and a well-run site operation will only trigger alerts exceptionally and the reporting regime is intended simply to provide an audit trail. In that sense it parallels a well-run health and safety policy. In the exceptional circumstance that an incident on site does result in damage to the church next door, the monitoring networks and audit trail would provide the evidence to support a claim.
A thoughtfully designed monitoring and alert system coupled with an actions protocol should slip seamlessly into the construction management plan that is now universally applied on all major development projects. Open communications and openly accessible data foster confidence on all sides. As a result, neighbourly relations will be enhanced and the risk of a catastrophic outcome reduced to the minimum.
