Investigating Bacterial Community Response During Discolouration Events in an Experimental Water Distribution System Cindy J. Smith, Rebecca L. Sharpe and Joby B. Boxall Pennine Water Group, Department of Civil and Structural Engineering, Sir Frederick

1
Potable water pipe-wall biofilm bacterial
community response to conditioning shear stress
and a hydraulic disturbance in a full-scale pipe
loop facility. C. J. Smith, R. Sharpe I.
Douterelo and J. B. Boxall School of
Natural Sciences, Microbiology, NUI Galway,
University Road, Galway, Ireland. Pennine Water
Group, Department of Civil and Structural
Engineering, Sir Frederick Mappin Building,
Mappin Street, University of Sheffield,
Sheffield, S1 3JD, UK. E-mail
cindy.smith_at_nuigalway.ie
The Problem Discolouration of potable water, due
to fine insoluble particles, is a major cause for
customer contacts to water companies. Within
Water Distribution Systems (WDS) pipe walls are
sites for biofilm development and the
accumulation of particulate material. The
stability and amount of material accumulated is
known to be influenced by the maximum shear
stress exerted by the daily flow profile but the
processes and mechanisms explicitly involved are
poorly understood. Mobilisation of material into
the bulk water occurs when shear stress exceeds
the conditioning values (Husband et al 2008). The
Hypothesis Biofilms play an important role in
understanding causes and consequences of material
layer at the pipe wall and hence in
discolouration. The Objective To examine the
effect of shear stress on biofilm bacteria
community development and the subsequent response
of WDS biofilm bacterial communities to increased
hydraulic conditions in a full scale WDS test
facility.
Figure 1 Pipe wall biofilm formation and
drinking water discoloration with increasing
shear stress above daily conditioning shear.
Experimental Set-up Methodology

• 3 x 200m pipe-loop test facility
• Biofilm material layers were accumulated for 28
  days at 8C under three different steady state
  boundary shear stresses 0.11, 0.22 and 0.44
  (N/m²).
• After 28 days, each loop was individually
  flushed according to figure 3. Each flushing step
  was conducted for three turnovers of water.
• Turbidity and DAPI cell counts were measured in
  the bulk water after three turnovers.
• Coupons were taken before and after the flushing
  event to analyze the bacterial community on the
  pipe wall.

Figure 3 Schematic of incremental shear stress
applied to each loop. Star indicates coupon
removal.

• DNA was extracted from coupon and the 16S rRNA
  gene amplified for T-RFLP, Q-PCR and gene
  sequencing.

Figure 2 The temperature controlled pipe loop
test facility. Insert the Pennine Water Group
coupon (Deines et al., 2010), 52 coupons are
inserted along the length of each loop to
facilitate examination of the pipe-wall biofilm.
Conditioning shear stress and mobilization of
pipe-wall material
Conditioning shear stress and biofilm community
structure
The effect of the mobilization event on pipe-wall
biofilm community structure
A Loop 1
C Loop 3
B Loop 2
B Turbidity (NTU) after each incremental
increase in shear
D Q-PCR of biofilm
2D Stress 0.07
Green line indicates 50 community similarity
based on Bray-Curtis similarity index.
Transform Square root
Figure 5 A) DAPI cell counts and B) turbidity in
drinking water after each incremental increase
in shear stress for loops 1, 2 3. The most
material was mobilized from loop 1, conditioned
at the lowest daily shear, as evidenced by the
increase in cell numbers and turbidity in the
drinking water.
Resemblance S17 Bray Curtis similarity
Coupon Pre-flush
Coupon Post-flush
Similarity ()
50
Figure 6 (A-C) MDS analysis of T-RFLP data from
loop 1, 2 3 before after mobilization. ANOSIM
analysis showed no difference in community
structure before and after the mobilization event
for any loop. (D) 16S rRNA gene copy numbers mm2
of pipe-wall before and after mobilization. No
statistical difference in gene copy numbers was
observed.

• Summary
• Conditioning shear stresses did not affect the
  bacterial community structure of a 28-day-old
  biofilm.
• Conditioning shear stress did affect mobilization
  of material into the bulk water - more material
  was mobilized by the lowest conditioning shear
  stress than the highest.
• Pipe wall biofilm community structure and 16S
  rRNA gene copy numbers were not altered by the
  mobilization event

Significance

• Results are moving us closer to an understanding
  of the link between daily conditioning shear,
  biofilm formation and discoloration.
• Flushing alone will not remove bacteria from WDS
  pipe-wall.