Abstract
Membrane bioreactors (MBR), used for wastewater treatment
in Ohio and elsewhere in the United States, have pore
sizes small enough to theoretically reduce concentrations of
protozoa and bacteria, but not viruses. Sampling for viruses
in wastewater is seldom done and not required. Instead,
the bacterial indicators Escherichia coli (E. coli) and fecal
coliforms are the required microbial measures of effluents for
wastewater-discharge permits. Information is needed on the
effectiveness of MBRs in removing human enteric viruses
from wastewaters, particularly as compared to conventional
wastewater treatment before and after disinfection.
A total of 73 regular and 28 quality-control (QC) samples
were collected at three MBR and two conventional wastewater
plants in Ohio during 23 regular and 3 QC sampling
trips in 2008–10. Samples were collected at various stages in
the treatment processes and analyzed for bacterial indicators
E. coli, fecal coliforms, and enterococci by membrane filtration;
somatic and F-specific coliphage by the single agar layer
(SAL) method; adenovirus, enterovirus, norovirus GI and GII,
rotavirus, and hepatitis A virus by molecular methods; and
viruses by cell culture. While addressing the main objective of
the study—comparing removal of viruses and bacterial indicators
in MBR and conventional plants—it was realized that
work was needed to identify data analysis and quantification
methods for interpreting enteric virus and QC data. Therefore,
methods for quantifying viruses, qualifying results, and applying
QC data to interpretations are described in this report.
During each regular sampling trip, samples were collected
(1) before conventional or MBR treatment (post-preliminary),
(2) after secondary or MBR treatment (post-secondary
or post-MBR), (3) after tertiary treatment (one conventional
plant only), and (4) after disinfection (post-disinfection).
Glass-wool fiber filtration was used to concentrate enteric
viruses from large volumes, and small volume grab samples
were collected for direct-plating analyses for bacterial indicators
and coliphage. After filtration, the viruses were eluted
from the filter and further concentrated. The final concentrated
sample volume (FCSV) was used for enteric virus analysis
by use of two methods—cell culture and a molecular method,
polymerase chain reaction (PCR). Quantitative PCR (qPCR)
for DNA viruses and quantitative reverse-transcriptase PCR
(qRT-PCR) for RNA viruses were used in this study.
To support data interpretations, the assay limit of detection
(ALOD) was set for each virus assay and used to determine
sample reporting limits (SRLs). For qPCR and qRT-PCR the
ALOD was an estimated value because it was not established
according to established method detection limit procedures.
The SRLs were different for each sample because effective
sample volumes (the volume of the original sample that was
actually used in each analysis) were different for each sample.
Effective sample volumes were much less than the original
sample volumes because of reductions from processing steps
and (or) from when dilutions were made to minimize the
effects from PCR-inhibiting substances. Codes were used to
further qualify the virus data and indicate the level of uncertainty
associated with each measurement.
Quality-control samples were used to support data interpretations.
Field and laboratory blanks for bacteria, coliphage,
and enteric viruses were all below detection, indicating that it
was unlikely that samples were contaminated from equipment
or processing procedures. The absolute value log differences
(AVLDs) between concurrent replicate pairs were calculated
to identify the variability associated with each measurement.
For bacterial indicators and coliphage, the AVLD results
indicated that concentrations ‹10 colony-forming units or
plaque-forming units per 100 mL can differ between replicates
by as much as 1 log, whereas higher concentrations can
differ by as much as 0.3 log. The AVLD results for viruses indicated
that differences between replicates can be as great as
1.2 log genomic copies per liter, regardless of the concentration
of virus. Relatively large differences in molecular results
for viruses between replicate pairs were likely due to lack of
precision for samples with small effective volumes.
Concentrations of E. coli, fecal coliforms, enterococci,
and somatic and F-specific coliphage in post-secondary and
post-tertiary samples in conventional plants were higher than
those in post-MBR samples. In post-MBR and post-secondary
samples, concentrations of somatic coliphage were higher than
F-specific coliphage. In post-disinfection samples from two
MBR plants (the third MBR plant had operational issues) and
the ultraviolet conventional plant, concentrations for all bacterial
indicators and coliphage were near or below detection;
from the chlorine conventional plant, concentrations in post-disinfection
samples were in the single or double digits. All of
the plants met the National Pollutant Discharge Elimination
System required effluent limits established for fecal coliforms.
Norovirus GII and hepatitis A virus were not detected
in any samples, and rotavirus was detected in one sample but
could not be quantified. Adenovirus was found in 100 percent,
enterovirus in over one-half, and norovirus GI in about one-half
of post-preliminary wastewater samples. Adenovirus and
enterovirus were detected throughout the treatment processes,
and norovirus GI was detected less often than the other two
enteric viruses. Culturable viruses were detected in post-preliminary
samples and in only two post-treatment samples from
the plant with operational issues.
U.S. Department of the Interior |
U.S. Geological Survey