Content uploaded by Christopher. P. Brough
Author content
All content in this area was uploaded by Christopher. P. Brough on Oct 13, 2023
Content may be subject to copyright.
Construction and Building Materials 408 (2023) 133404
Available online 12 October 2023
0950-0618/© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
High risk concrete blocks from County Donegal: The geology of defective
aggregate and the wider implications
C. Brough
*
, B. Staniforth, C. Garner , R. Garside , R. Colville , J. Strongman , J. Fletcher
Petrolab Limited, C Edwards Ofces, Gweal Pawl, Redruth, Cornwall TR15 3AE, UK
ARTICLE INFO
Keywords:
Free mica
Pyrrhotite
Defective concrete
IS465
Donegal
ABSTRACT
Since 2019, geologists at Petrolab have tested concrete block samples from over 1800 properties in County
Donegal and observed deterioration patterns based on the aggregate types used within those concrete blocks.
Across these properties more than 25 aggregate types have been used within concrete block, with seven distinct
types considered high risk when assessed against relevant standards. With respect to those considered high risk,
elevated free mica and elevated reactive sulphides (predominantly pyrrhotite) contents appear to play a critical
role, particularly in phyllite aggregates commonly found in defective blockwork. The particular concentration of
both of these mineral groups is considered the main cause of the rapid degradation often observed with this
aggregate type. Other high-risk aggregates have lower amounts of one or both of these mineral groups, but they
are still present in concentrations that are cause for concern and associated in some instances with damaged
property.
These ndings have important implications. Firstly, with two mineral groups to dene risk, the extent of risk
can be split depending on the concentrations of the different minerals with potentially different remedial
pathways followed in each instance. Aggregates with high concentrations of both free mica and sulphides
(predominantly pyrrhotite) will pose the highest risk, likely requiring the most signicant intervention in the
long-term, while aggregates with only free mica and low sulphide content will be lower risk. In support of this,
none of the properties tested to date, where only free mica risk is present within the concrete samples, have gone
on to display the most serious degradation category. This, coupled with the evidence of internal sulphate attack
present within the most high-risk pyrrhotite-bearing aggregate, is further evidence of the importance of internal
sulphate attack in the highest risk aggregates. For all seven aggregates of concern the authors provide general
recommendations for remedial measures, or further research based on the ndings of this study.
Secondly, the risk window for affected properties appears wider than initially believed. It spans for at least two
decades with a signicant window encompassing structures built between 1995 and 2010. The currently
recorded latest occurrence is 2015, but it is not clear from the current data that this latest occurrence reects the
last use of defective concrete block due to the inherent latency period between building construction and the
time for degradation to set-in. Consequently, a long-term testing regime is necessary to provide condence in
property conveyancing within County Donegal for properties built within, and since, the dened risk window.
1. Introduction
The County Donegal Mica crisis rose to prominence in 2011–2013
when homeowners in County Donegal, Ireland, began to notice severe
cracking and structural damage to their homes. Following investigation
by an expert committee established by the then Minister of Housing and
Urban Renewal, it was suspected that the principal cause of the damage
was due to the use of defective concrete blocks, which had deteriorated
over time, causing the blocks to crumble and weaken the structures [24].
The issue was compounded by the fact that many of the affected
homeowners did not have the means to pay for the necessary repairs,
which cost well in excess of
€
100,000 per property. The crisis led to
widespread public outcry and calls for action from the Irish government,
national regulatory bodies and local authorities. In 2019, a government
grant scheme was established, which has since been further scaled up-
wards to help affected homeowners cover the cost of repairs, with the
government committing several billion Euros so far to the fund.
The issue is particularly prevalent in homes built during the 1990s
* Corresponding author at: Petrolab Ltd, UK.
E-mail address: chris@petrolab.co.uk (C. Brough).
Contents lists available at ScienceDirect
Construction and Building Materials
journal homepage: www.elsevier.com/locate/conbuildmat
https://doi.org/10.1016/j.conbuildmat.2023.133404
Received 6 July 2023; Received in revised form 15 September 2023; Accepted 16 September 2023
Construction and Building Materials 408 (2023) 133404
2
and 2000s, with an estimated 5,000 properties affected in County
Donegal (e.g. [24]). The cause of the blockwork deterioration was
initially interpreted to be due to elevated levels of disseminated free
mica in the binder [24]. It was posed in a subsequent conference paper
discussing the defective concrete in County Donegal that the presence of
elevated free mica within the binder was increasing the microporosity of
the cement matrix and thereby increasing susceptibility to secondary
processes such as frost or leaching, and reducing the overall strength
[11]. Previous publications had indicated mica concentrations within
the binder of mass concrete of just 1 % could negatively impact concrete
performance [12,18,17], and an original technical report published in
1963 had raised serious concerns about incremental increases in mica
content [9]. However, the original government research in 2017 and
subsequent guidance did not directly address the additional issues
related to the presence of high concentrations of iron sulphide miner-
alisation, specically pyrrhotite, within the affected blockwork. Exces-
sive quantities of pyrrhotite and associated fracturing began to be
routinely identied within microscopic analysis. Recent research has
conrmed the presence of elevated levels of pyrrhotite present within
aggregate in concentrations which exceed current EN 12620 compliance
values [5,16], and these authors make the case that this is producing
internal sulphate attack and this is the primary cause of deterioration
largely unrelated to the free mica content [16].
The evidence of recent analysis is that the high-risk aggregates in
widespread use, including the widely implicated pyrrhotite-bearing
phyllite, have been locally sourced from County Donegal. It is there-
fore important to lay out the geology of the County, which has been
studied and well understood for a little over a century now (e.g.
[29,1,19,20]).
County Donegal is located in the northwest of Ireland and can be
summarised as having formed during ve distinct periods of geological
activity (Fig. 1):
(i) The rst of these is the Rhinns Complex, a Paleoproterozoic
basement which underlays much of North Donegal and consists of
Precambrian orthogneiss. This is only rarely observed at the
surface, being seen at the Northern edge of the Inishown Penin-
sula on the island of Inishtrathull [25] and at the Tor rocks off
Fig. 1. Geological map of Donegal (Data from Geological Survey of Ireland, 2017), with inset from [22] showing the extent of the Dalradian Supergroup across the
Republic of Ireland and the UK.
C. Brough et al.
Construction and Building Materials 408 (2023) 133404
3
Malin Head. These outcrops date to around 1.8 billion years ago
and consist mainly of gneiss, schist, and quartzite.
(ii) The second main geological period originates from the Dalradian
Supergroup which was lain down during the Neoproterozoic and
forms the majority of the metasedimentary formations in County
Donegal [13]. Lithologically the Dalradian Supergroup shows
stratigraphic and lateral continuity with similar metasediments
in Scotland and Northern Ireland which have been well docu-
mented (e.g. [29] – Ireland & [2] for Scotland) (Fig. 1). Across
this lateral continuity the Dalradian is split into four groups
(Grampian Group, Appin Group, Argyll Group and Southern
Highland Group – Fig. 1), of which three can be observed in
County Donegal (Fig. 1). The Dalradian Supergroup was origi-
nally deposited as a series of mudstones and sandstones, some
variably calcareous, and including rarer limestones. Many of the
mac intrusions observed throughout the North of Donegal were
intruded prior to or during the early phases of the Caledonian
orogeny [15].
(iii) Thirdly, the Dalradian Supergroup has been impacted by the
Grampian Phase of the Caledonian orogeny and there are igneous
intrusions (granites) related to this orogeny that can be found
throughout County Donegal, most notably in the main Donegal
granite (e.g. [29,19]). These intrusions were emplaced around
420–390 million years ago. The orogeny also folded, faulted, and
metamorphosed the Dalradian Supergroup into the present range
of quartzite, pelitic schist, semi-pelitic and psammitic schists,
phyllites and marbles seen throughout the metasedimentary
sequence. Occasional occurrences of higher metamorphic grade
metasediments are noted with the rare occurrence of amphibo-
lites. Garnets are also variably present as an indicator mineral.
(iv) Fourthly, there are minor occurrences of Devonian old red
sandstones, and Carboniferous sedimentary units, primarily to
the South of Donegal, with some rare outcrops on the Northeast
along the coast. The Carboniferous units include sandstones,
siltstones, shales, conglomerates and limestones [19].
(v) Finally, there are important Paleogene features throughout
County Donegal, some of which have also produced aggregate in
common use. This includes the impact of glaciation which
extensively reshaped the landscape and left behind deposits of
glacial till and boulder clay [23]. Rivers have also eroded and
reworked the major geological units leading to the widespread
deposition of sands and gravels.
2. Materials and methods
In 2019, the Irish Government rst introduced a grant scheme for
affected homeowners that provides nancial assistance to remediate
their home. This scheme requires homes to be analysed to conrm they
have a problem with defective concrete blocks before they can access the
scheme [26]. Over the last four years geologists at Petrolab have ana-
lysed concrete blocks from >2500 properties across the Republic of
Ireland of which approximately 1800 properties have been within
County Donegal (Fig. 2).
The analysis has followed the procedures outlined in I.S. 465 relating
Fig. 2. A point density heat map showing the locations of 2016 properties tested across the Republic of Ireland. The heat map includes the locations of all tested
properties regardless of nal risk categorisation, but if plotting for just high-risk the distribution would look the same at this resolution. The plot was made in python
using the Plotly graphing libraries. The base map is provided by OpenStreetMap [28].
C. Brough et al.
Construction and Building Materials 408 (2023) 133404
4
to the assessment and testing of concrete blocks from damaged prop-
erties [26]. Samples are provided by I.S. 465 registered engineers,
typically from Building Condition Assessment (BCA) Group 3 or Group 4
classied properties (I.S. 465 Table 1, “Damaged” and “Signicantly
damaged” respectively) with geologists at Petrolab acting as the pro-
fessional geologist. As part of the I.S. 465 standard there are four Groups
relating to damage severity with Group 4 being the most damaged
(Table 1). When testing in accordance with the I.S. 465 standard for
application to the original redress scheme at least 8 samples would be
collected covering every wall type, inclusive of inner leaf, outer leaf and
below DPC concrete. This process breaks down the testing into three
stages (Suite A - Macroscopic, Suite B – Microscopic, chemical testing,
physical testing and risk assessment and Suite C – Additional testing).
In addition to this testing regime a centralised database has been
kept retaining the key results of each test which has been utilised to look
at overall patterns for defective concrete within County Donegal ranging
from the aggregate in use through to the year of build. In the context of
the overall results, it is noted that sampling has generally focussed on
damaged properties rather than a random sampling of properties within
County Donegal as a whole. Properties tested for the original Defective
Concrete Blocks Grant Scheme (DCBS) account for 85 % of the analysed
samples, with property tested for conveyancing adding the remaining
samples. Sampling within the DCBS scheme is very rigid and geared
towards sampling inner leaf, outer leaf and below damp proof course
(DPC), and at least initially was not focussed on sampling any specic
damaged portions of the property. As such a disproportionate amount of
initial results were focussed on the examination of currently sound inner
leaf samples of properties displaying group 3 or 4 damage in the
associated outer leaf blockwork.
At Suite A the submitted samples were examined as received using a
Nikon SMZ-U stereoscopic microscope with bre optic illuminator. An
aggregate type is assigned to each concrete sample and a preliminary
risk categorisation given. Photographs of the received cores, including a
cut surface of the main concrete type are taken and recorded. Risk at
Suite A is assigned based on the visual “soundness“ of the core sample,
the proportion of problematic lithologies (sulphide- or mica-bearing)
and visual evidence of aggregate or cement paste deterioration. If the
concrete is progressed to Suite B testing, a thin section (45 mm ×30 mm,
30 µm thick) and accompanying polished block is prepared from each
selected sample using standard water-sensitive geomaterials sample
preparation techniques. These sections are examined by conventional
transmitted, reected uorescent and reected light polarising micro-
scopy using a Nikon or Zeiss research microscope. Results from this
testing are then reported to the client and compiled into the overall
database. A three-letter code is given to the sample representing the
geological composition of the aggregate that has been used (e.g. PHY =
phyllite aggregate). This three-letter code can be used to look for pat-
terns by aggregate type that are deducible across all the completed tests.
As part of testing of defective concrete blockwork in County Donegal
additional testing from I.S. 465 Test Suite C is also undertaken. To satisfy
the requirement to analyse free mica content of the blockwork, Petrolab
have developed a simplied methodology to quantify liberated musco-
vite mica <63 µm as a function of the cement paste via SEM/EDX
analysis. “Free mica” is dened within the I.S. 465 standard as mono-
mineralic muscovite mica occurring as discrete or ‘free’ aky grains,
which can adversely affect the nal strength and durability of the
hardened concrete. No value is given within the standard but pre-
liminary observations by Eden and Vickery [11] suggest that the per-
formance of the concrete was adversely affected when free mica content
was greater than 5 %. This assessment of the free mica content of the
binder is undertaken on the same 45 mm ×30 mm epoxy resin vacuum
impregnated sub-sample of the selected specimen used for thin section
preparation. This sample ‘chip’ is polished and carbon coated prior to
SEM examination. Three areas of interest (AOI) are selected on the
cement paste at 500x magnications to represent typical binder
composition in each sample. Coarse aggregate particles, “locked”
muscovite mica and signicant voids are manually excluded. Each AOI
is analysed using a ZEISS EVO MA 25 scanning electron microscope
(SEM) located at Petrolab Ltd, Redruth and tted with two Bruker xFlash
6|60 x-ray detectors for energy-dispersive X-ray spectroscopy (EDX)
analysis. Images and EDX phase maps of the AOI are acquired using
Bruker Esprit software 2.1. Image post-processing and area measure-
ments are carried out using Fiji/ImageJ 1.47 image analysis software
and plugins [32], utilising threshold analysis of the individual elemental
maps in combination with the manual exclusions to express “free mica”
as a proportion of cement paste.
Assessment of the degradation risk inherent to each concrete type (i.
e. each three-letter aggregate code) has been made, where possible, by
analysing two principal risk factors, namely the abundance of free mica
in the binder as per the requirement of I.S. 465 and the abundance of
primary sulphides (particularly pyrrhotite) and compliance against the
criteria given in EN 12620 and S.R. 16. In addition to this, attention was
paid to the presence of BCA reports conrming the presence of damage
to the property (i.e. a BCA Group 3 or 4). Chemical analysis was un-
dertaken on material taken adjacent to the select thin/polished block
section locations. Analysis was carried out by a UKAS accredited labo-
ratory in accordance with B.S. EN 1744-1 [6], clauses 10, 11.2 and 12.
The analysis procedure involves high temperature combustion in an
atmosphere of pure oxygen, followed by quantication by infrared de-
tectors. Quantication limits are to the nearest 0.01 %. Corrections were
made to the reported total sulphur (TS) data to account for S that will be
naturally present as part of the cement (taken to be ~ 0.1 % as
mentioned in Annex E of I.S. 465) and to account for the density dif-
ference between binder and aggregate. This latter value results in a
Table 1
Building grouping levels of damage (reproduced from I.S. 465).
Group Damage Building Condition Assessment
Group
1
Undamaged Pattern cracking is not present, however some or all
the circumstantial evidence is recorded in the
Chartered Engineer’s Report
Group
2
Damaged Pattern cracking is present in at least one elevation
(but insufcient evidence of other damage to
classify the building as Group 4, see Group 4, a) to
e)), and no circumstantial evidence
a
is recorded in
the Chartered Engineer’s Report
Group
3
Damaged Pattern cracking is present in at least one elevation
(but insufcient evidence of other damage to
classify the building as Group 4, see Group 4, a) to
e)), and some or all the circumstantial evidence
a
is
recorded in the Chartered Engineer’s Report
Group
4
Signicantly
damaged
Pattern cracking on at least one elevation, and at
least two of the following further items of damage
present on same or adjacent elevation:
Vertical cracks near corners >5 mm in width;
Crumbling concrete blocks;
Severe displacement of reveals with cracking;
Wall leaning or bulging noticeably i.e. local
deviation of slope in the horizontal or vertical plane
of external walls of >1 in 100, and
Cracking of widths >1 mm on internal leaf where
damage is also present on the corresponding
external leaf, or multiple cracks of concrete masonry
walls in one room of >0.5 mm
Where circumstantial evidence is available it shall
be recorded in the Chartered Engineer’s Report.
a
Circumstantial evidence (risk factors) suggesting the possible presence of deleterious
materials in concrete blocks includes:
Information that blocks came from manufacturer(s) reported to have supplied
blocks to other damaged dwellings likely to have arisen from deleterious material in
concrete block,
Construction within the date range of constructions mentioned in the Report of the
Expert Panel on Concrete Blocks [24], and in the geographic areas reported to be
affected; and
Documented information (e.g. Chartered Engineer’s Report) that other dwellings in
the same estate or locale have exhibited signs of damage likely to have arisen from
deleterious material in concrete blocks.
C. Brough et al.
Construction and Building Materials 408 (2023) 133404
5
multiplier of 1.08 to the resultant TS value (e.g. [16]).
3. Results
The results of testing have identied at least 25 distinct concrete
types in use within County Donegal. The average free mica percentage
has been assessed for each of these and there are ve which contain
average free mica concentrations in excess of 5 %. The 25 concrete types
were also assessed for pyrrhotite content and there were ve which
contained average pyrrhotite concentrations potentially in breach of
compliance values. Some of these aggregate types were common to both
sets (i.e. contained both elevated free mica and elevated pyrrhotite
concentrations). There has also been rare observation of a deleterious
limestone aggregate similar to that seen in County Mayo concrete blocks
and deleterious limestone ll identied in multiple counties across
Ireland (e.g. [21,8]. These aggregates of concern can be summarised
using three-letter codes along with their aggregate descriptions
(Table 2). Along with these aggregates of concern a description is also
provided for the most common stable aggregate observed throughout
County Donegal with generally lower free mica and very low to absent
pyrrhotite (RGV, Table 2).
The most common of the high-risk concretes is the phyllite-bearing
aggregate (PHY & PHQ). This concrete type is dominated by crushed
0/20 mm phyllite (variably metamorphosed mudstones and siltstones)
and micaceous quartzite (metamorphosed impure sandstones, including
irregular sized quartz and alkali feldspar clasts) (Fig. 3). PHY and PHQ
can be considered as two subsets of the same essential aggregate source,
one where phyllite is a major component (PHY) and one where phyllite
is a minor component (PHQ). Both aggregates also contain minor
calcareous micaceous quartzite fragments, which have probably been
derived from metamorphosed calcareous sandstones. These rocks have
not been sufciently metamorphosed to form schist, as the “mica”
crystals (interlayered muscovite and chlorite) are too ne grained to be
visible to the naked eye. The presence of chlorite and absence of biotite
are also indicative of a fairly low grade of metamorphism. Much of the
phyllite aggregate contains a prominent crenulation cleavage typical of
polyphase deformation at low metamorphic grade. Free mica is
commonly dispersed through the binder in very high (>10 vol% of
cement) concentrations. This free mica has the same composition as the
mica within the aggregate, and is interpreted to have been abraded from
the aggregate during manufacture due to the softness of the phyllite in
particular. The aggregate (particularly phyllite) contains appreciable
amounts of pyrrhotite, with additional pyrite and occasional observa-
tion of chalcopyrite, marcasite and cobaltite. Pyrrhotite often shows
evidence of oxidation, with the characteristic iron-oxide striations,
particularly of liberated grains in the binder, or on the edge of the
aggregate (Fig. 4). Some of this oxidation likely represents stockpile
alteration prior to manufacture whilst some of it will be in-situ, partic-
ularly where there is evidence of binder staining or alteration adjacent
to the pyrrhotite.
The PGV aggregate is dominated by a 0/16 mm gravel containing
highly micaceous components (typically well rounded phyllite, with
minor pelitic schist, quartzite and some igneous components) (Fig. 3).
The well-rounded nature of the aggregate components, occasional
presence of extraneous material (e.g. wood) and low sulphide levels due
to weathering in the natural environment conrm this as a gravel
source. In some cases this aggregate is of Low/Medium risk, whilst in a
subset of cases there are very high (>10 vol%) levels of free mica in the
binder which are interpreted to have been derived from the phyllite-rich
fragments.
The WMQ aggregate is dominated by a 0/20 mm weathered mica-
ceous quartzite and weathered greywackes, with high concentrations of
Fe oxides and Fe oxyhydroxides (goethite and ferrihydrite) still adhered
to the aggregate which can be seen by the ochre-coloured orange frag-
ments in hand specimen (Fig. 3) or as light grey anhedral masses in
reected light (Fig. 4). Petrographic analysis has also observed some
rare relics of primary sulphates, either preserved with the aggregate, or
implied from sulphates found in the binder. The aggregate is commonly
observed mixed with a ne sand (river, beach or quartz), and when this
is present the free mica proportion tends to be lower. Free mica is
commonly dispersed through the binder in high (>5 vol% of cement
paste) concentrations. This free mica has the same composition as the
mica within the aggregate, and is interpreted to have been abraded from
the aggregate during manufacture due to the softness of the weathered
micaceous components in particular. Sulphide concentrations within the
aggregate are very low having been naturally weathered to iron oxides
within the depositional environment.
The MSD code aggregate is dominated by crushed 0/20 mm slate and
weakly metamorphosed mudstones and siltstones (metapelite) and
partly cleaved chloritised micaceous quartzite (metamorphosed impure
sandstones, including metamorphosed greywacke with irregular sized
quartz clasts) (Fig. 3). The rocks are less metamorphosed than the
typical deleterious phyllite blocks from County Donegal, with a ner
muscovite grain size, and also lack the crenulation cleavage observed in
the PHY & PHQ aggregate. Free mica is dispersed through the binder in
high (>5 vol%) concentrations. This free mica has the same composition
as the mica within the aggregate, and is interpreted to have been
abraded from the aggregate during manufacture due to the softness of
the metapelite components in particular. Sulphide mineralisation within
this aggregate is typically pyrite dominant with pyrrhotite subordinate
and occasionally absent. Oxidation of the pyrite is observed (Fig. 4),
though this frequently appears to be from stockpile alteration. This
aggregate is usually considered high risk due to the proportion of free
mica and abundance of reactive sulphides.
The SCH code aggregate has been rarely observed within properties
from County Donegal but is sufciently distinct geologically to be
considered a separate lithology. This concrete is dominated by crushed
Table 2
Summary aggregate codes of the high-risk concrete types identied in County
Donegal, along with the proportion of high risk aggregate they represent. Also
included is the stable aggregate RGV for comparison.
Code
1
Aggregate Components Typical I.S. 465
risk as reported
Number
2
% of
High
risk
PHY Phyllite, micaceous
quartzite, minor vein
quartz
High 1219 86
PGV Phyllite-rich river gravel High 67 5
PHQ Micaceous quartzite, minor
phyllite
High 41 3
WMQ Weathered micaceous
quartzite & greywacke
High 39 3
MSD Low-grade metasediments,
slates, siltstones,
High 30 2
SCH Schist, marble, quartzite,
amphibolite
High 16 1
BGL Argillaceous limestone,
dolomitised limestone
High 12 1
HFS
3
Pelitic hornfels, gabbro,
metapelite
Low/Medium 39 –
RGV Heterogeneous
metasedimentary gravel
Low/Medium 246 –
1
These are summary codes for concrete containing aggregates of a specic
mineralogy. Occasionally these aggregates have also been mixed with a second sub-
ordinate aggregate (e.g. a ne beach sand). Where this second aggregate is clearly
subordinate in nature and stable it has been grouped in with the dataset of the
dominant aggregate.
2
n =1709, and this refers to the number of concrete samples tested to Suite B, and
included in this dataset. This is not the same as the total number of properties as
multiple concrete types may be present in the same property and may often have been
tested.
3
Although a typically Low/Medium risk aggregate type this has been included
within the dataset due to concerns over the pyrrhotite content which is just in breach of
compliance on average (see Table 3).
C. Brough et al.
Construction and Building Materials 408 (2023) 133404
6
0/20 mm biotite or muscovite pelitic to psammitic schist (+/−garnet)
and quartzite (metamorphosed sandstones) (Fig. 3). It also commonly
contains marbles, likely intercalated with the schist within the original
geological setting, and some rare amphibolite fragments. Free mica is
dispersed through the binder in low (<2%) to high (>5%) concentra-
tions. This free mica has the same composition as the mica within the
aggregate, and is interpreted to have been abraded from the aggregate
during manufacture. As with the phyllite aggregate pyrrhotite is
commonly present with the hosting aggregate primarily associated with
the pelitic schist and intercalated marble, and shows evidence of
oxidation, either during stockpiling or during in situ reaction.
The BGL aggregate contains a crushed 0/16 mm limestone. This
limestone shows different textures in general from the typical black
argillaceous limestone/calcareous mudstone encountered within
County Mayo during similar I.S. 465 testing and appears to be a slightly
different lithology. It is sourced from the Carboniferous units to the
South of the County of Donegal (Fig. 1), as the Dalradian metasediments
were deposited before the onset of shelly life. By comparison with the
Mayo aggregate it contains a signicant proportion of sparry dolomite
and more common “clean” bioclastic limestone with lower clay contents
in the matrix (Fig. 3). However, it still contains a dark grey/black
argillaceous limestone component with a texture made of local clusters
of stylolites, as well as locally common laminated calcareous mudstone.
These stylolites contain locally common clays, organic carbonaceous
material, and framboidal pyrite and often form sub-parallel laminations.
The framboids are 5 – 20
μ
m across, with individual microcrystals <2
Fig. 3. Examples of High-risk aggregates and RGV aggregate observed in County Donegal (A:PHY/PHQ, B:PGV, C:WMQ, D:BGL, E:SCH, F:HFS, G:MSD & H:RGV.
Scale at top left, samples are 103 mm diameter and nominal 100 mm thick.
C. Brough et al.
Construction and Building Materials 408 (2023) 133404
7
μ
m. Oxidation in degraded samples is observed, either as characteristic
“daisy” alteration where the framboids form a tight cluster, or as indi-
vidual microcrystal alteration where the clusters are looser (Fig. 4).
Framboidal pyrite present within calcareous mudstone and laminated
argillaceous limestone is known to be reactive and has been implicated
as a source of sulphate in previous cases of pyritic heave within Ireland
[21] and in concrete degradation (e.g. [3]). This aggregate is nearly
always High risk when encountered within County Donegal.
The HFS aggregate is dominated by crushed 0/15 mm metasedi-
mentary and metabasic igneous rocks including metagabbro, meta-
dolerite, quartz-biotite hornfels, spotted hornfels, amphibolite and
metapelite (Fig. 3). These materials likely represent a combination of
mac igneous intrusions (e.g. dolerite) and thermally metamorphosed
pelitic country rock surrounding the intrusion and were probably
sourced from the same quarry. The igneous components show wide-
spread hydrothermal alteration and replacement by various minerals
including chlorite and epidote. Pyrrhotite and pyrite are typically
observed within the biotite hornfels component of the aggregate. The
concentrations are generally low and the aggregate is usually considered
Low/Medium risk, but occasionally there may be locally abundant
concentrations of pyrrhotite. The pyrrhotite within this aggregate is
generally unreacted or only weakly reacted. This likely reects its pre-
dominant host within nely crystalline hornfels, which may act as an
effective protection to oxidation. The free mica content is always <5 %.
A pertinent observation is that the HFS aggregate is observed either as
an all-in aggregate or a second subordinate aggregate mixed in with a
primary crushed white quartzite (WHT). Where present as a subordinate
aggregate it has typically been crushed to 0/4 mm, and despite being
present in lower proportions (~approximately 30 % of overall aggre-
gate) has been locally observed to contain pyrrhotite and has been
observed with damaged buildings. This combined aggregate (WHT +
HFS) isn’t considered further within this paper as more research is
required.
The RGV aggregate is dominated by a variable 0/15 mm gravel
containing a diverse range of sedimentary, metasedimentary and
igneous components without any one dominant lithology. Typical
components include schist, quartzite, granite, dolerite, phyllite, sand-
stones and chert (Fig. 3). The well-rounded nature of the aggregate
components, occasional presence of extraneous material (e.g. wood) and
low sulphide levels due to prior weathering conrm this as a gravel
source. In nearly all cases this aggregate is of Low/Medium risk.
Nonetheless, in a very rare subset of cases there are very high levels of
free mica in the binder.
Nearly all of the aggregates with elevated free mica and/or elevated
pyrrhotite show association with building condition assessment reports
rated either Group 3 or 4 (i.e. damaged to severely damaged, I.S. 465,
Table 3), and contain either excessive free mica (>5%), sulphide con-
centrations above compliance values, sulphide within problematic li-
thologies or a combination of these (Table 3, Fig. 5, Fig. 6). It is noted
that not all property containing high-risk aggregate shows evidence of
degradation at the time it was tested (see Table 3).
With respect to pyrrhotite grades the concentrations for the different
pyrrhotite-bearing aggregates can be plotted up and compared to the
criteria given in EN 12620 (Fig. 6), and all pyrrhotite-bearing aggregates
exceed this compliance value. It is noted that for most of these aggre-
gates pyrrhotite and pyrite are both present as sulphide within the
aggregate and pyrrhotite is a major component of the sulphide balance.
The exception to this is the MSD aggregate where pyrrhotite is present in
subordinate concentrations to pyrite, usually signicantly so and to the
point where it is sometimes absent altogether. Nonetheless, the straight
Fig. 4. Examples of sulphide and oxide mineralisation encountered within the primary High-Risk aggregates in County Donegal (A:pyrrhotite within aggregate and
oxidised pyrrhotite within binder, from PHY; B:Oxidised pyrite on the edge of an aggregate fragment, from MSD; C:Fine framboidal (fr) and cubic (cu) pyrite with
oxidation rims from BGL aggregate; D:Fe oxides from WMQ.
C. Brough et al.
Construction and Building Materials 408 (2023) 133404
8
reading of EN 12620 refers only to aggregate that contains pyrrhotite
regardless of whether the pyrrhotite is the dominant sulphide or not.
By cross referencing the aggregate in use with the year of build as
derived from the BCA report, a risk window for the use of defective
concrete block can be shown (Fig. 7). The main risk window runs from
1992 up until at least 2015 with a peak occurring in the late 1990s and
dropping off sharply in 2008. The risk window extends for longer than
the original window highlighted in the expert report [24], starting both
earlier and with no certainty on the dened nish point yet, although
the timing of that report in 2017 may act as a constraint on the use of
phyllite block beyond that. It is also notable from the analysis that over a
quarter of properties tested to the I.S. 465 scheme contain multiple
concrete types (Fig. 8). The presence of multiple concrete types within a
property is a complicating factor when considering the cause of damage
within a property, particularly if multiple different high-risk aggregates
are present within the same property (cf. Table 3). This particularly
impacts some of the WMQ & rare PGV aggregates which are occasionally
found in the same property as the PHY aggregate, and where PHY is the
more dominant concrete block observed. Multiple concrete types within
the same property may be due to building extensions, and while this may
be the case in some instances it is also observed that there is commonly a
difference between below DPC and above DPC concrete, suggesting that
this particular difference originates from the time of construction.
4. Discussion
The cause of deterioration within concrete blocks within County
Donegal has been the source of signicant debate since it began
appearing in homes in increasing numbers. In 2017 a desktop review
appointed by the government concluded that excessive free mica was
likely the cause making blocks vulnerable to wet-dry cycling and
freeze–thaw damage [24]. A conference paper published by Eden and
Vickery [11], observed that in most cases of poor concrete durability the
free mica content was >5 %, and this value gained traction within the
wider public. Furthermore, the role of free mica has been known as a
deleterious component within concrete for decades [9,12,18,17] and for
that reason the hypothesis of excessive mica as the principal cause of
deterioration was considered plausible and initially widely accepted.
Since then, evidence of deterioration relating to pyrrhotite has been
assessed and evidence of deterioration attributed to sulphide oxidation
Table 3
Average free mica and total sulphur (TS) contents of the aggregates summarised in Table 2, along with the number of different BCA reports associated with each
aggregate type.
Code Number Average Free Mica (vol%) Average TS content (wt%) No BCA received BCA 1 BCA 2 BCA 3 BCA 4
PHY 1219 12 0.43 818 3 3 269 126
PGV 67 8 0.08 51 1 0 9 4*
PHQ 41 11 0.23 26 0 0 14 1
WMQ 39 9 0.02 28 0 1 8 1*
MSD 30 7 0.26 21 0 0 8 1
SCH 16 5 0.36 13 0 0 1 2
BGL 12 – 0.63** 6 0 0 5 1
HFS 39 3 0.15 38 0 0 1 0
RGV 246 4 0.10** 216 1 1 23 4*
*
These BCA 4 values for WMQ, PGV and RGV are observed in properties that also contain PHY or BGL concrete. The same is true for all except 3 of the BCA 3 values for RGV.
**
In both BGL and RGV the only sulphide typically present is pyrite. For BGL this sulphur is hosted in framboidal pyrite in deleterious aggregate and is a cause for concern. For
RGV this sulphur is hosted in cubic pyrite and as such the levels are signicantly below the applicable EN 12620 guideline value, whilst also being within tolerance of the lower
pyrrhotite standard on the rare occasions that pyrrhotite is observed within the RGV aggregate.
Fig. 5. Box and whisker plots showing the range of ‘free mica’ as a proportion of cement for different high-risk aggregate types containing free mica, along with RGV
for comparison. N.B. BGL is not included as there is no free mica risk, and concentrations have not been measured. Total number used for each dataset is as per
Table 2, as each test had Suite C free mica analysis, with the exception of MSD where n =29. Shown with a red dashed line is the value of free mica set out in Eden &
Vickery, 2019 where they have observed that in most cases of poor concrete block durability the free mica content of the paste is typically >5 % by volume of the
paste. (For interpretation of the references to colour in this gure legend, the reader is referred to the web version of this article.)
C. Brough et al.
Construction and Building Materials 408 (2023) 133404
9
have been identied by Petrolab and other research organisations
[4,16]. Pyrrhotite or reactive pyrite have been recognised as being
present within the most problematic aggregates associated with the
most severe and rapid cases of concrete degradation, at concentrations
that exceed compliance values and with degraded concrete containing
evidence of internal sulphate attack [16]. Within that paper evidence
was also presented that the climatic conditions for years have not met
the conditions required for freeze–thaw cycling, effectively ruling that
out as a principal cause for the deterioration.
With the evidence presented for internal sulphate attack within
phyllite block [16] it is important then to consider whether free mica in
the binder still plays a role in the deterioration observed. Whilst it is
known that excessive free mica within the binder causes issues during
concrete manufacture, and as a nal product, with several case studies
and technical reports illustrating this [9,12,18,30,17,34,27], all studies
have examined the effect of free mica within traditional mass concrete
applications rather than in extruded precast masonry products with low
cement content and high voidage. Only the study by Leemann & Holzer
examined the effects of micas at different size fractions. Therefore, the
question remains as to the material signicance, if any, it is playing in
the defective concrete of County Donegal. Within concrete assessed as
part of this study there were two concrete types observed which contains
excessive free mica without sulphide concentrations (WMQ & PGV) and
these can be used to investigate the hypothesis that free mica on its own
Fig. 6. Box and whisker plots showing the range of total sulphur concentrations for different high-risk aggregate types containing pyrrhotite, and also showing RGV
for comparison. (n =90 for PHY, n =4 for PHQ, n =2 for PGV, n =1 for WMQ, n =7 for SCH, n =15 for HFS, n =3 for MSD, n =10 for RGV). The red line shows
the compliance value for aggregate containing pyrrhotite from EN 12620 (Note: in most instances this does not apply to the RGV aggregate, as pyrrhotite is usually
absent, but even when it does RGV is compliant). (For interpretation of the references to colour in this gure legend, the reader is referred to the web version of
this article.)
Fig. 7. Histogram showing the main risk window for tested properties containing high-risk concrete within County Donegal, and some key dates within that range.
C. Brough et al.
Construction and Building Materials 408 (2023) 133404
10
is causing degradation. Within this study a common observation,
particularly within the WMQ aggregate, is of excessively high micro-
porosity binder, with increased capillary porosity suggestive of a high to
very high water-cement ratio during concrete preparation. The resultant
binder is highly porous and often contains examples of poor binder to
aggregate adhesion (Fig. 9A, B & C), with rare primary sulphates also
present in the binder, and extensive Fe oxide and Fe oxyhydroxides
(goethite) adhered to fragment surfaces (e.g. Fig. 4D). The example
illustrated in Fig. 9B was taken from the outer leaf of the property. The
below DPC concrete from the same property, which was the same
concrete type, was severely degraded to the point of being unsound.
Binder with a high capillary porosity and poor binder to aggregate
adhesion will be more susceptible to a number of secondary processes,
examples of which would include freeze–thaw cycling, wet-dry cycling,
external sulphate attack and internal sulphate attack. It is noted that for
high-risk aggregate containing only elevated free mica as a risk (i.e.
WMQ or PGV), and where this was the only aggregate observed within
the property, damage has to date only been recorded to BCA Group 3.
There have been no critically deteriorated examples of WMQ analysed at
Suite B, but at Suite A the most frequently degraded samples received
Fig. 8. Pie chart showing the number of concrete types per property as identied at Test Suite A within County Donegal, when tested under the conditions of the
original Defective Concrete Blocks (DCBS) scheme. Based on the testing of 1643 properties.
Fig. 9. Examples of high microporosity zones around the margins of aggregate fragments within WMQ (A, C & D) & PGV (B) concrete, particularly in images A
(uorescent light), B (cross polarised light) & C (transmitted light). Image B shows alignment of micas along the aggregate binder interface. The below DPC concrete
from the same property was severely fragmented (unsound). Image D shows an example of a fracture running through binder within WMQ concrete.
C. Brough et al.
Construction and Building Materials 408 (2023) 133404
11
has been from below DPC concrete which is likely to relate to the wet-
dry cycling and loss of cohesion at the binder-aggregate interface due
to high microporosity zones (Fig. 9). For all associations of these ag-
gregates with a BCA Group 4 property, PHY, SCH or BGL was also pre-
sent within the same property, and in every instance bar one PHY, SCH
or BGL was the more common concrete observed. This suggests that free
mica is likely a deleterious factor in its own right, but that the combi-
nation of free mica and sulphides is required to lead to BCA Group 4, and
the severe degradation widely reported. It is important to note that PHY
concrete also contains high microporosity binder for the same reasons as
WMQ, namely the disseminated ne mica that has resulted in high
capillary microporosity. For that reason, the PHY is generally associated
with poor quality binder, analogous to that seen in WMQ and PGV
concrete.
With respect to iron sulphides, the presence of elevated concentra-
tions of these (in particular pyrrhotite) within concrete is known to be
problematic and has been implicated in the premature degradation of
concrete in numerous case studies and theoretical studies (e.g.
[7,31,33,35]), particularly so in Eastern Connecticut, Massachusetts and
Trois-Rivi`
eres in Quebec, Canada. (e.g. [14,10]). The recent publication
by Leemann et al., [16] specically showed the example of severe in-
ternal sulphate attack within four examples of critically degraded
properties from County Donegal. From properties assessed during this
study numerous examples of in-situ oxidation and degradation have been
Fig. 10. Examples of degradation associated with pyrrhotite oxidation in phyllite (PHY) aggregate from different samples. A: Photomicrograph in plane-polarised
light showing an example of fracturing network through the binder and radiating from sulphides in the phyllite; B: Reected photomicrograph showing heavily
oxidised pyrrhotite spalling into the binder. C: Photomicrograph in plane-polarised light showing large sulphide grain with oxidation staining the surrounding binder
and evidence of fracturing associated with the adjacent binder. D: Photomicrograph in plane polarised light showing internal fracturing running along foliation and
intersecting with oxidised pyrrhotite grains which also run parallel to foliation. E; Photomicrograph in plane polarised light showing sulphide fracturing and spalling
into the adjacent void. F: Examples of gypsum replacing heavily leached binder and gypsum induced spalling of heavily oxidised pyrrhotite.
C. Brough et al.
Construction and Building Materials 408 (2023) 133404
12
observed, with at least 44 critically deteriorated samples and many more
showing localised examples of internal sulphate attack. At the beginning
of testing in accordance with I.S. 465 such examples were by no means
prevalent due to the focus on cores collected from intact portions of the
inner leaf of the property. As testing has progressed more examples have
been observed particularly as the focus shifted towards the importance
of pyrrhotite. Petrographic observation of damage typically presents as
pyrrhotite oxidation, with iron-staining of surrounding binder (see
image 10C) and spalling of aggregate, pyrrhotite or binder into voids
(image 10B & 10E) (Fig. 10). The nal fracture pattern within the
concrete may often present as an interlinked network running through
aggregate and binder (image 10A).
An important mineralogical feature of the pyrrhotite mineralisation
within the phyllite aggregate is that it usually runs parallel to the foli-
ation (see image 10D), which means that oxidation of pyrrhotite is likely
to lead to preferential cracking along the foliation plane as a zone of
weakness. This is also a problem that is intrinsic to new or currently
unreacted block, as the manufacture of concrete blocks sometimes in-
duces internal cracks which terminate at contact with the binder, indi-
cating they formed early and are unrelated to degradation at this stage.
This is nonetheless a problem as these early formed cracks will often
intersect pyrrhotite within the phyllite meaning that oxidation is more
likely in the future. In severe cases pyrrhotite oxidation has been asso-
ciated with secondary sulphate formation in the cement paste, particu-
larly ettringite and thaumasite, but also gypsum in fully carbonated
samples (see image 10F), with subsequent weakening of the cement
paste that is typical of internal sulphate attack.
With the presence of two mineral groups as risk factors a risk graph
can be constructed that assesses high risk concrete against both these
mineral groups (Fig. 11). Within this chart the further an aggregate type
goes to the top right beyond both indicated red lines the higher the
degree of risk. From the chart it is apparent that phyllite aggregate
(PHY) is the highest risk aggregate on both variables and this is likely to
be a primary driver for the speed of degradation observed within con-
crete blocks with PHY aggregate. For aggregate in the lower left corner
(i.e. RGV) the risk of degradation from aggregate related problems can
be considered negligible to low.
Nonetheless it is also noted that concrete with lower free mica levels
and equivalent sulphide concentrations (SCH) has shown signs of
degradation, and that concrete with lower total sulphur and lower free
mica has evidence of associated degradation (e.g. PHQ aggregates). For
WMQ there is only one datapoint at present for total sulphur, as it is very
rarely tested for this and sulphide concentrations are very low. Petro-
graphically, it has been observed to contain rare primary sulphates
(gypsum) that may be observed in the binder or as part of the aggregate.
These primary sulphates are probably derived from sulphides original to
the aggregate that have weathered out within the geological environ-
ment. The extent of weathering observed within the WMQ aggregate and
frequent evidence of sub-rounding of aggregate fragments suggests a
source within the glacial till/glacial uvial deposits within County
Donegal, and the presence of Fe oxides (goethite), sulphates and high
free mica are considered the primary constituents of concern reducing
aggregate to binder adhesion (e.g. Fig. 9), and increasing susceptibility
to secondary processes. MSD aggregate is clearly of elevated risk within
this scheme but there is uncertainty over the degree of risk as pyrrhotite
is usually only a trace component of the total sulphide deportment, and
the pyrite content, if considered alone, would be within the EN 12620
guideline value. The free mica is also not as elevated as that observed in
WMQ, PGV, PHQ or PHY. It is recommended that further work on this
aggregate is undertaken to clarify the long-term risks. HFS has been
included within this study due to the presence of pyrrhotite and the
technical breach of EN 12620 guidelines. At present there is no petro-
graphic evidence for deterioration observed with this aggregate and
only one building with a BCA Group 3. Nevertheless, given the technical
breach of a standard further work is required to understand the long-
term risks associated with concrete built with this aggregate.
The other aggregate noted in Table 2 & 3 is BGL. The BGL aggregate
degradation mechanism relates to oxidation of framboidal pyrite within
calcareous mudstones to argillaceous limestones leading to gypsum
production, spalling of aggregate (Fig. 12) and eventual cracking of the
binder [3].
As properties containing defective concrete can be expected to
degrade at different rates depending on external factors (e.g., driving
rain index, ground conditions) and internal factors (e.g., the presence
and abundance of free mica and sulphides) there will be a variable rate
of degradation observed both between different aggregate types and
Fig. 11. A risk rating chart showing the average free mica in the binder against total sulphur level for the seven aggregates that contain mica and may contain
pyrrhotite in Table 3. The n values for each point are recorded in Figs. 5 & 6. The error bars refer to the interquartile range where n >10 (i.e. contain 50 % of the data
range). Error bars are only shown for TS values where at least 10 data points are available. Also shown are the allowable TS limit for aggregates containing pyrrhotite
(EN 12620) and the free mica content from Eden and Vickery [11].
C. Brough et al.
Construction and Building Materials 408 (2023) 133404
13
within aggregate types. Nonetheless, concrete made with phyllite
aggregate containing both free mica and reactive sulphides (i.e. pyr-
rhotite) at the highest concentrations has to date been associated with
the highest number of damaged properties (see Table 3), along with the
most severely damaged properties (i.e. the number of BCA Group 4
properties) and this correlation is considered signicant.
What is unclear from this analysis, is the extent to which the other
high-risk aggregates have been in use within County Donegal and
whether the lower proportion they represent of the high-risk aggregates
is a function of lower levels of use, or a slower speed of degradation
which manifests as being encountered less frequently within the current
testing regime. The presence of multiple concrete types within proper-
ties is an important feature to note as future testing of property,
particularly for conveyancing will need to be aware that below DPC
concrete, inner leaf concrete and outer leaf concrete may contain
different concrete types, and in some properties, it has even been
observed that there have been multiple types of concrete used within the
outer leaf. Where risk varies by aggregate type, it is important when
sampling to minimise the chance of missing concrete within properties
which may be high risk. As such there is now a necessity to formalise a
long-term testing regime for properties not currently displaying damage
that are to be bought and sold within the conveyancing market, which
were themselves constructed between ~ 1990 and ~ 2020. The testing
regime used for the conveyancing market needs to provide certainty to
buyers and lenders that their purchase is safe and give homeowners
more clarity on how they can sell or, if necessary, remediate their
property. This testing regime must not only collect a sufcient number
of samples, most likely along a similar process to the original DCBS or
RICS scheme in Cornwall, but must also test for all the considered risks
(i.e. mica, pyrrhotite and pyrite). Failure to do so will place unaccept-
able risk on homeowners, buyers and lenders.
5. Conclusion
Since 2019 geologists at Petrolab have tested in excess of 1800
properties from County Donegal and have noted correlations of concrete
deterioration by aggregate type that point to a critical role for both
elevated free mica and pyrrhotite contents which carry several impli-
cations. The rst of these is that the phyllite aggregate which makes up
the overwhelming majority of defective blockwork in use contains both
mineral groups present at the highest concentrations and this is
considered to be the primary cause of the speed of degradation.
The second implication is that with two mineral groups to dene risk
that high-risk concrete could be split on the extent to which both of these
variables are present, and potentially different remediation pathways
recommended. Across the 1800 properties tested, there hasn’t yet been
one property containing concrete with excessive free mica and low
sulphide content that has recorded a BCA Group 4 (i.e. signicantly
damaged). This strongly suggests that for aggregates with just excessive
free mica (i.e. WMQ and PGV) that a remediation pathway may be
possible without demolition, either allowing for inner leaf retention,
and/or outer leaf retention if the building envelope is robust, similar to
the original scope of the Defective Block Grant Scheme. The principal
risk in the presence of elevated free mica may relate to the below DPC
concrete in this instance, likely to receive the greatest wet-dry cycling
ux. For concrete containing both elevated free mica and elevated
pyrrhotite (PHY, PHQ & SCH), the evidence from this study is that the
free mica has resulted in the production of high microporosity binder,
and that subsequent internal sulphate attack on this porous binder has
led to severe degradation eventually resulting in BCA Group 4 houses.
The remediation pathway in this instance will likely be restricted to
demolition and rebuild. For high-risk BGL concrete a similar pathway
will likely be the only option. For the remaining two aggregates, MSD
requires further research and clarication on the risk when pyrrhotite is
present in trace amounts and pyrite is the dominant sulphide. HFS
aggregate has very few associated instances of building damage and no
petrographic evidence yet observed of deterioration. It nonetheless is
often in technical breach of the EN 12620 standard and further research
will be required on the long-term risk of this aggregate type and whether
remediation will be required.
The third implication is that the risk window for affected properties
(i.e. the year of rst use of defective blockwork up until the last year)
would be wider than initially thought, with a core window which lasts
for ~ 15 years (1995–2010) and a long-term window which may reach
three decades, including past 2010. There is no dened closure to the
window yet, though it is likely given the observation of defective con-
crete in the early 2010s, and publications within that decade, that poor
quality aggregate was discontinued or signicantly reduced from use
within that decade. This leads to a nal implication which is that for
properties built within a dened risk window there will need to be a
long-term testing regime set up for property conveyancing within
County Donegal from properties built at least within ~ 1990 and ~
2020. This testing regime needs to collect a sufcient number of sam-
ples, possibly along a similar vein to the Defective Concrete Block Grant
Scheme, and test for all known risks (i.e. mica, pyrrhotite and pyrite) to
provide condence to homeowners, buyers and lenders that their pur-
chase is safe.
CRediT authorship contribution statement
C. Brough: Validation, Supervision, Project administration, Inves-
tigation, Formal analysis, Conceptualization, Writing - original draft,
Fig. 12. Example of gypsum formation within concrete blocks containing BGL aggregate within County Donegal. Most gypsum is space lling (i.e. void linings) but
there is evidence of gypsum induced spalling to the top-right of image A and to bottom centre of image B.
C. Brough et al.
Construction and Building Materials 408 (2023) 133404
14
Writing - review & editing. B. Staniforth: Methodology, Validation,
Investigation, Writing – review & editing, Formal analysis. C. Garner:
Methodology, Investigation, Conceptualization, Supervision. R. Gar-
side: Writing – review & editing, Validation, Investigation. R. Colville:
Methodology, Data curation, Visualization. J. Strongman: Conceptu-
alization, Methodology, Project administration, Supervision. J.
Fletcher: Conceptualization, Data curation, Supervision, Writing - re-
view & editing.
Declaration of Competing Interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgements
The authors would like to thank three detailed reviews and the
feedback of the Editor as critical to the nal paper. Former and current
colleagues at Petrolab Ltd are also thanked for numerous clarifying
dicussions around the analysed defective concrete. Rebecca Bolton is
thanked for her assistance with Figure 1. The authors would also like to
thank the I.S. 465 Engineers who have frequently provided further in-
formation and contextulisation of results for individual tests.
References
[1] G.I. Alsop, D.H.W. Hutton, A review and revision of Dalradian stratigraphy in
central and southern Donegal, Ireland, Irish J. Earth Sci. 10 (1990) 181–198.
[2] R. Anderton, Sedimentation and tectonics in the Scottish Dalradian, Scott. J. Geol.
21 (1985) 407–436.
[3] A. Bromley, A compendium of mundic aggregate used in SW England, 2002.
[4] C. Brough, B. Staniforth, C. Garner, J. Strongman, J. Fletcher, R. Colville, The
Donegal Mica Scandal, the Tip of an Iceberg?, No. EGU23-967, Copernicus
Meetings, 2023.
[5] BS E.N. 12620: 2002+A1: 2008. Aggregates for Concrete. British Standards
Institution, London.
[6] BS E.N. 1744-1:2009+A1. (2012). Tests for Chemical Properties of Aggregates
Chemical Analysis. British Standards Institute, London.
[7] I. Casanova, L. Agull´
o, A. Aguado, Aggregate expansivity due to sulde
oxidation—I. Reaction system and rate model, Cem. Concr. Res. 26 (7) (1996)
993–998.
[8] M.A. Czerewko, J.C. Cripps, Investigation of destructive ground heave attributed to
pyritic ll affecting new-build properties in the Dublin area of Ireland, Eng. Geol.
299 (2022), 106534.
[9] J.D. Dewar, ‘Effects of mica in the ne aggregate on the water requirement and
strength of concrete’, TRA/370, Cement and Concrete Association, London, 1963.
[10] J. Duchesne, A. Rodrigues, B. Fournier, Concrete damage due to oxidation of
pyrrhotite-bearing aggregate: a review, RILEM Techn. Lett. 6 (2021) 82–92.
[11] Eden, M & Vickery, Sandberg LLP. (2019) Investigating the causes of deterioration
in concrete blocks in Southern Ireland. 17th EMABM, University of Toronto,
Toronto, Canada, May 20-23.
[12] P.G. Fookes, W.A. Revie, Mica in concrete – a case history from Eastern Nepal,
Concrete 16 (3) (1982) 12–16.
[13] Harris, A. L., Baldwin, C. T., Bradbury, H. J., Johnson, H. D. & Smith, R. A. (1978).
Ensialic sedimentation: the Dalradian Supergroup. In: Crustal Evolution in
Northwest Britain and adjacent regions (edited by Bowes, D. R. & Leake, B. E.).
Special Report 6. Geological Society, London, 52-75.
[14] D. Jana, Pyrrhotite epidemic in Eastern Connecticut: diagnosis and prevention, ACI
Mater. J. 117 (1) (2020) 1–20.
[15] B.E. Leake, The metabasites and garnet amphibolites of Glencolumbkille, Co.,
Donegal and the early mac intrusions into the Dalradian rocks of Donegal,
Connemara and Scotland, Irish J. Earth Sci. 34 (2016) 27–43, https://doi.org/
10.3318/ijes.2016.34.27.
[16] A. Leemann, B. Lothenbach, B. Münch, T. Campbell, P. Dunlop, The “mica crisis” in
Donegal, Ireland–A case of internal sulfate attack? Cem. Concr. Res. 168 (2023),
107149.
[17] Leemann, A & Holzer, L. (2001) Inuence of mica on the properties of mortar and
concrete, Proceedings of the 8
th
Euroseminar on Microscopy Applied to Building
Materials, pp 199 – 204.
[18] T.P. Lees, Impurities in concreting aggregates, British Cement Association Ref. 45
(1987) 016.
[19] Long, C. B., & B. J. McConnell. (1997) Geology of north Donegal: A geological
description to accompany the bedrock geology 1: 100,000 scale map series, Sheet 1
and part of Sheet 2, north Donegal. Geological Survey of Ireland.
[20] Long, CB, McConnell, BJ, Alsop, GI, O’Connor, PI, Claringbold, K & Cronin, C
(1999), Geology of South Donegal: A geological description of South Donegal, to
accompany the Bedrock Geology 1:100,000 Scale Map Series, Sheet 3 and part of
Sheet 4, South Donegal. Geological map series of Ireland 1:100,000, no. Sheet 3
and part of sheet 4, Geological Survey of Ireland.
[21] G.D. Matheson, P. Quigley, Evaluating pyrite-induced swelling in Dublin
mudrocks, Q. J. Eng. Geol. Hydrogeol. 49 (1) (2016) 47–66.
[22] C.A. McAteer, S.J. Daly, M.J. Flowerdew, M.J. Whitehouse, C.L. Kirkland,
A Laurentian provenance for the Dalradian rocks of north Mayo, Ireland, and
evidence for an original basement–cover contact with the underlying Annagh
Gneiss Complex, J. Geol. Soc. Lond. 167 (2010) 1033–1048.
[23] McCabe, M. (2007). Glacial geology and geomorphology. The landscapes of
Ireland. Publisher: Dunedin, Edinburgh ISBN: 978-1-90376-587-6.
[24] D. McCarthy, N. Kane, F. Lee, D. Blaney, Report of the Expert Panel on Concrete
Blocks, Department of Housing, Local Government and Heritage, 2017 https://
www.gov.ie/en/publication/0218f-report-of-the-expert-panel-on-concrete-blocks/
.
[25] R.J. Muir, W.R. Fitches, A.J. Maltman, The Rhinns Complex: Proterozoic basement
on Islay and Colonsay, Inner Hebrides, Scotland, and on Inishtrahull, NW Ireland,
Earth Environ. Sci. Trans. R. Soc. Edinb. 85 (1) (1994) 77–90.
[26] National Standards Institute of Ireland. I.S. 465:2018+A1:2020. Assessment,
testing and categorisation of damaged buildings incorporating concrete blocks
containing certain deleterious materials and Amendment 1:2020.
[27] E. Novikova, H.L. Arvidsson, K.-J. Miskovsky, The behaviour of mica rich base
course aggregates temperature climate conditions, Luleå, University of
Technology, ISSN 1402–175 (2008) (2008) 25.
[28] OpenStreetMap contributors. (2015). Retrieved from https://planet.
openstreetmap.org.
[29] W.S. Pitcher, A.R. Berger, The geology of Donegal: a study of granite emplacement
and unroong, Wiley Interscience, New York, 1972, p. 435.
[30] J.W. Schmitt, Effects of Mica, Aggregate Coatings, and Water-Soluble Impurities on
Concrete. Volume 12, Issue number 12 (1990). American Concrete Institute ISSN:
0162–4075.
[31] T. Schmidt, A. Leemann, E. Gallucci, K. Scrivener, Physical and microstructural
aspects of iron sulde degradation in concrete, Cem. Concr. Res. 41 (3) (2011)
263–269.
[32] J. Schindelin, I. Arganda-Carreras, E. Frise, V. Kaynig, M. Longair, T. Pietzsch,
S. Preibisch, C. Rueden, S. Saalfeld, B. Schmid, J.-Y. Tinevez, D. White,
V. Hartenstein, K. Eliceiri, P. Tomancak, A. Cardona, Fiji: an open-source platform
for biological-image analysis, Nat. Methods 9 (7) (2012) 676–682. Versatile Wand
plugin tool by Michael Schmid.
[33] A. Tagnit-Hamou, M. Saric-Coric, P. Rivard, Internal deterioration of concrete by
the oxidation of pyrrhotitic aggregates, Cem. Concr. Res. 35 (1) (2005) 99–107.
[34] Y. Wakizaka, K. Ichikawa, Y. Nakamura, S. Anan, Deterioration of concrete due to
specic minerals, Environ. Econ. 2 (2001) 331–338.
[35] H. Xueqin, C. Pang, R. Gao, S. Wu, S. Sun, Iron sulde-related internal sulfate
attack: microstructure and mechanism, Mag. Concr. Res. 70 (8) (2018) 379–389.
C. Brough et al.