The paper presents the findings from field studies of fly ash concrete
structures containing reactive (alkali-silica) aggregates. Data is pr
esented from a number of hydraulic structures in Wales and Ontario con
structed using geologically similar greywacke-argillite aggregates. Al
l the structures with ash showed evidence of damage due to ASR; indeed
reaction with this aggregate has lead to replacement of structures in
both areas. Reaction may occur at alkali levels significantly below 3
kg/m(3) Na2Oe and has been found in a structure with a measured 'avai
lable' alkali content of less than 2 kg/m(3) (NaOe)-O-2. However, the
fly ash concrete structures are in excellent condition after more than
25 years service despite having higher alkali contents than many of t
he damaged structures; the Lower Notch Dam was actually constructed us
ing a high alkali cement (> 1.0% Na2Oe). Microstructural and pore solu
tion analysis showed that much of the alkali in the fly ash concrete w
as 'bound' in the CSH and not available to the pore solution for ASR.
Published findings related to the McPherson Test Road, the cracking ob
served in fly ash concrete being attributed to drying shrinkage. Detai
ls regarding the use of ash in the Hanshin Expressway are not adequate
ly reported, however, it would appear that 18% fly ash was not effecti
ve in completely suppressing expansion in this structure. This is not
surprising in view of the moderate level of ash and excessive alkali c
ontent of the concrete (up to 0.33% Na2Oe). The balance of field perfo
rmance data sustains the concept of using Class F fly ash to control e
xpansion due to ASR in concrete, provided the material is used at suff
icient levels of replacement. Further control of the alkalis from othe
r sources would appear to be a prudent precaution even when the concre
te contains fly ash. It is surprising that many specifications have ad
opted the advice from accelerated laboratory tests and penalize fly as
h by assigning an 'alkali contribution' (e.g. one-sixth the total alka
li or all of the ASTM C311 'available' alkali) to the material which m
ay often preclude its use. Such advice appears to contradict the more
cogent argument from the field where Class F fly ash has been successf
ully used for many decades with no reported incidences of ASR in struc
tures containing sufficient levels of ash.