I am writing to express our practice's concern over the article 'Explaining the bookend effect' by Clive Richardson in aj 16.7.98.
The 'bookend effect' is his unproven theory about terraced houses. It is not an established behaviour for such buildings and the 'remedy' he proposes for it is likely to do more harm than good to the properties involved.
Alan Baxter and Associates has worked on the repair and alteration of over 1000 London terraced houses and has not found evidence of a bookend effect, nor do we know of any other engineer who acknowledges it. Mr Richardson's article itself does not give any evidence which supports his claim that progressive expansion movements are taking place.
Georgian and Victorian brick-and-timber terraced houses move around for many reasons, the principal ones being:
heave or settlement of foundations - often triggered by late alterations, drainage works or the effects of trees
movements of timbers built into the brickwork - whether due to moisture changes or decay of the timber
differential movements within the brickwork itself due to carbonation of lime mortar, different brick stresses internally and externally, or chemical changes in the bricks or the mortar.
Flank walls at the ends of terraces have a particular propensity to lateral movements because there is little restraint from adjoining walls or floors. Indeed, they usually have a staircase next to them which prevents the floors being connected to them for much of their length. So flank walls often bulge or lean outward by considerable amounts, the movements normally having occurred quite soon after they were built. Terraces with major ground-floor openings in their front, rear and spine walls generally experience more significant movements because there is less lateral restraint.
Mr Richardson only quotes one set of measurements in support of his theory, for a terrace in Baker Street. For his theory to hold, there should be progressive leaning-out of each party wall towards the ends of the terrace. This is not the case - one of the flank walls is 148mm out of plumb, but the party wall next to it is only 17mm out. The flank wall at the other end of the terrace is measured at 27mm out of plumb, but such a figure is little more than the variation one would expect across the surface of a wall built of the hand-made bricks of the period. None of the examples Mr Richardson cites shows any progressive expansion movement. This is most likely because no such movement is occurring.
There are two reasons we are concerned lest Mr Richardson's theory is left unchallenged:
The needless worry caused to all those who might think that the flank walls of all terraces will eventually fall over unless significant structural works are undertaken
Mr Richardson's proposed 'remedy' of cutting movement joints at the party wall lines of the houses would tend to weaken the integrity of their structures and allow further movements to occur due to the reduced restraint caused by the joint.
Where terraced houses exhibit signs of movement, each case is best assessed carefully in its own context, so that appropriate remedies (if any are needed) may be adopted. Such remedies are more likely to involve additional local ties to help knit the structure of a terrace together rather than the cutting-in of joints to create concentrated lines of weakness.
Alan Baxter and Associates