Skip to contents

Usage (Uniform Inflow)

Michael Rustler

08 June, 2022

Source: vignettes/usage.Rmd
usage.Rmd

How to work with the kwb.qmra package in R(Studio) is described in the following chapters

Once the R package is installed it can be loaded with the following command in R(Studio):

1 Input data

1.1 Location of ‘dummy’ configuration

The folder with the csv configuration files for a hypothetical use case is located here:

#### DEFINE DIRECTORY ################
confDir <- system.file("extdata/configs/dummy", package = "kwb.qmra")
confDir
## [1] "/Users/runner/work/_temp/Library/kwb.qmra/extdata/configs/dummy"

The following screenshot shows the required configuration files located in confDir (here: /Users/runner/work/_temp/Library/kwb.qmra/extdata/configs/dummy):

Screenshot of required configuration files

1.2 Import configuration into R

All csv files with the input data for the hypothetical u dummy (as shown above) are imported into R with the following function:

#### LOAD ############################
config <- config_read(confDir)

2 Check input data

The QMRA will be performed - in case the user does not modify them in R - based on the imported input data, which are defined in the configuration folder.

In case of the dummy configuration, a Monte carlo simulation (n = 10) for 365 exposure events per year for three pathogens and the following input parameters will be performed:

Simulated pathogens for QMRA (defined in: ‘inflow.csv’ with simulated = 1):
PathogenID PathogenName PathogenGroup
3 Campylobacter jejuni and Campylobacter coli Bacteria
32 Rotavirus Viruses
36 Giardia duodenalis Protozoa
Inflow concentrations (defined in: ‘inflow.csv’) for pathogens used for QMRA:
PathogenID PathogenName PathogenGroup type min max
3 Campylobacter jejuni and Campylobacter coli Bacteria uniform 10 10000
32 Rotavirus Viruses uniform 10 10000
36 Giardia duodenalis Protozoa uniform 10 10000
Treatment schemes (defined in: ‘treatment_schemes.csv’) used for QMRA:
TreatmentSchemeID TreatmentSchemeName TreatmentID TreatmentName
1 Berlin (BF + Slow sand) 8 Slow sand filtration
1 Berlin (BF + Slow sand) 9 Bank filtration
2 Depth & surface filtration 1 Conventional clarification
2 Depth & surface filtration 5 Granular high-rate filtration
2 Depth & surface filtration 15 UV
Treatment processes (defined in: ‘treatment_proecesses.csv’) and assumed log-reductions used for QMRA (from WHO, 2011):
TreatmentID TreatmentName TreatmentGroup PathogenGroup type min max
1 Conventional clarification Coagulation, flocculation and sedimentation Bacteria uniform 0.20 2.0
1 Conventional clarification Coagulation, flocculation and sedimentation Protozoa uniform 1.00 2.0
1 Conventional clarification Coagulation, flocculation and sedimentation Viruses uniform 0.10 3.4
5 Granular high-rate filtration Filtration Bacteria uniform 0.20 4.4
5 Granular high-rate filtration Filtration Protozoa uniform 0.40 3.3
5 Granular high-rate filtration Filtration Viruses uniform 0.00 3.5
8 Slow sand filtration Filtration Bacteria uniform 2.00 6.0
8 Slow sand filtration Filtration Protozoa uniform 0.30 5.0
8 Slow sand filtration Filtration Viruses uniform 0.25 4.0
9 Bank filtration Pretreatment Bacteria uniform 2.00 6.0
9 Bank filtration Pretreatment Protozoa uniform 1.00 2.0
9 Bank filtration Pretreatment Viruses uniform 2.10 8.3
15 UV Primary disinfection Bacteria uniform 4.00 4.0
15 UV Primary disinfection Protozoa uniform 4.00 4.0
15 UV Primary disinfection Viruses uniform 4.00 4.0
Ingested volume per event (defined in: row 3 ‘volume_perEvent’ in ‘exposure.csv’) (source: own assumption)):
name type min max mode
volume_perEvent triangle 0.5 3 1.5
Dose-response models (defined in: ‘doseresponse.csv’) used for QMRA (from QMRAwiki):
PathogenID PathogenName PathogenGroup Best fit model* k alpha N50 Host type Dose units Route Response Reference Link
3 Campylobacter jejuni and Campylobacter coli Bacteria beta-Poisson NA 0.144 890.00 human CFU oral (in milk) infection Black et al 1988 http://qmrawiki.canr.msu.edu/index.php/Campylobacter_jejuni_and_Campylobacter_coli:_Dose_Response_Models
32 Rotavirus Viruses beta-Poisson NA 0.253 6.17 human FFU oral infection Ward et al, 1986 http://qmrawiki.canr.msu.edu/index.php/Rotavirus:_Dose_Response_Models
36 Giardia duodenalis Protozoa exponential 0.0199 NA NA human Cysts oral infection Rendtorff 1954 http://qmrawiki.canr.msu.edu/index.php/Giardia_duodenalis:_Dose_Response_Models
Health parameters (defined in: ‘health.csv’) for simulated pathogens (from WHO, 2011):
PathogenID PathogenName infection_to_illness dalys_per_case
3 Campylobacter jejuni and Campylobacter coli 0.70 0.0046
32 Rotavirus 0.03 0.0140
36 Giardia duodenalis 0.30 0.0015

4 Run risk calculation

Subsequently the risk calculation can be performed in R(Studio) by executing the following code, which uses the config that was imported and inspected above:

risk <- kwb.qmra::simulate_risk(config)
## 
## # STEP 0: BASIC CONFIGURATION
## 
## Simulated 3 pathogen(s): Campylobacter jejuni and Campylobacter coli, Rotavirus, Giardia duodenalis
## Number of random distribution repeatings: 10
## Number of exposure events: 365
## 
## # STEP 1: INFLOW
## 
## Simulated pathogen: Campylobacter jejuni and Campylobacter coli
## Create 10 random distribution(s): uniform (n: 365, min: 10.000000, max: 10000.000000)
## Simulated pathogen: Rotavirus
## Create 10 random distribution(s): uniform (n: 365, min: 10.000000, max: 10000.000000)
## Simulated pathogen: Giardia duodenalis
## Create 10 random distribution(s): uniform (n: 365, min: 10.000000, max: 10000.000000)
## Providing inflow events ... ok. (0.00s) 
## Providing inflow paras ... ok. (0.00s) 
## 
## # STEP 2: TREATMENT SCHEMES
## 
## Create 10 random distribution(s): uniform (n: 365, min: 0.200000, max: 2.000000)
## Create 10 random distribution(s): uniform (n: 365, min: 1.000000, max: 2.000000)
## Create 10 random distribution(s): uniform (n: 365, min: 0.100000, max: 3.400000)
## Create 10 random distribution(s): uniform (n: 365, min: 0.200000, max: 4.400000)
## Create 10 random distribution(s): uniform (n: 365, min: 0.400000, max: 3.300000)
## Create 10 random distribution(s): uniform (n: 365, min: 0.000000, max: 3.500000)
## Create 10 random distribution(s): uniform (n: 365, min: 2.000000, max: 6.000000)
## Create 10 random distribution(s): uniform (n: 365, min: 0.300000, max: 5.000000)
## Create 10 random distribution(s): uniform (n: 365, min: 0.250000, max: 4.000000)
## Create 10 random distribution(s): uniform (n: 365, min: 2.000000, max: 6.000000)
## Create 10 random distribution(s): uniform (n: 365, min: 1.000000, max: 2.000000)
## Create 10 random distribution(s): uniform (n: 365, min: 2.100000, max: 8.300000)
## Create 10 random distribution(s): uniform (n: 365, min: 4.000000, max: 4.000000)
## Create 10 random distribution(s): uniform (n: 365, min: 4.000000, max: 4.000000)
## Create 10 random distribution(s): uniform (n: 365, min: 4.000000, max: 4.000000)
## Simulated treatment: Conventional clarification for Bacteria
## Simulated treatment: Conventional clarification for Protozoa
## Simulated treatment: Conventional clarification for Viruses
## Simulated treatment: Granular high-rate filtration for Bacteria
## Simulated treatment: Granular high-rate filtration for Protozoa
## Simulated treatment: Granular high-rate filtration for Viruses
## Simulated treatment: Slow sand filtration for Bacteria
## Simulated treatment: Slow sand filtration for Protozoa
## Simulated treatment: Slow sand filtration for Viruses
## Simulated treatment: Bank filtration for Bacteria
## Simulated treatment: Bank filtration for Protozoa
## Simulated treatment: Bank filtration for Viruses
## Simulated treatment: UV for Bacteria
## Simulated treatment: UV for Protozoa
## Simulated treatment: UV for Viruses
## Simulated treatment: Conventional clarification for Bacteria
## Simulated treatment: Conventional clarification for Protozoa
## Simulated treatment: Conventional clarification for Viruses
## Simulated treatment: Granular high-rate filtration for Bacteria
## Simulated treatment: Granular high-rate filtration for Protozoa
## Simulated treatment: Granular high-rate filtration for Viruses
## Simulated treatment: Slow sand filtration for Bacteria
## Simulated treatment: Slow sand filtration for Protozoa
## Simulated treatment: Slow sand filtration for Viruses
## Simulated treatment: Bank filtration for Bacteria
## Simulated treatment: Bank filtration for Protozoa
## Simulated treatment: Bank filtration for Viruses
## Simulated treatment: UV for Bacteria
## Simulated treatment: UV for Protozoa
## Simulated treatment: UV for Viruses
## Joining, by = "TreatmentID"
## Joining, by = "TreatmentID"
## 
## # STEP 3: EXPOSURE
## 
## Simulated exposure: volume per event
## Create 10 random distribution(s): triangle (n: 365, min: 0.500000, max: 3.000000, mode = 1.500000)
## Joining, by = c("PathogenGroup", "eventID", "repeatID")
## Joining, by = c("eventID", "repeatID")
## 
## # STEP 4: DOSE RESPONSE
## 
## # A tibble: 3 × 13
##   PathogenID PathogenName   PathogenGroup `Best fit mode…`       k  alpha    N50
##        <dbl> <chr>          <chr>         <chr>              <dbl>  <dbl>  <dbl>
## 1          3 Campylobacter… Bacteria      beta-Poisson     NA       0.144 890   
## 2         32 Rotavirus      Viruses       beta-Poisson     NA       0.253   6.17
## 3         36 Giardia duode… Protozoa      exponential       0.0199 NA      NA   
## # … with 6 more variables: `Host type` <chr>, `Dose units` <chr>, Route <chr>,
## #   Response <chr>, Reference <chr>, Link <chr>
## 
## # STEP 5: HEALTH
## 
## # A tibble: 3 × 4
##   PathogenID PathogenName                        infection_to_il… dalys_per_case
##        <dbl> <chr>                                          <dbl>          <dbl>
## 1          3 Campylobacter jejuni and Campyloba…             0.7          0.0046
## 2         32 Rotavirus                                       0.03         0.014 
## 3         36 Giardia duodenalis                              0.3          0.0015
## Joining, by = c("PathogenID", "PathogenName")

All (input & output) data will be saved in the resulting R object risk which an be easily inspected by the user, e.g.:

Input data

str(risk$input, 1)
## List of 5
##  $ inflow      :List of 1
##  $ treatment   :List of 1
##  $ exposure    :List of 1
##  $ doseresponse:List of 1
##  $ health      : tibble [3 × 4] (S3: tbl_df/tbl/data.frame)

Output data

str(risk$output, 1)
## List of 4
##  $ events          : tibble [54,750 × 20] (S3: tbl_df/tbl/data.frame)
##  $ total           : grouped_df [60 × 15] (S3: grouped_df/tbl_df/tbl/data.frame)
##   ..- attr(*, "groups")= tibble [60 × 6] (S3: tbl_df/tbl/data.frame)
##   .. ..- attr(*, ".drop")= logi TRUE
##  $ stats_total     : grouped_df [54 × 14] (S3: grouped_df/tbl_df/tbl/data.frame)
##   ..- attr(*, "groups")= tibble [6 × 6] (S3: tbl_df/tbl/data.frame)
##   .. ..- attr(*, ".drop")= logi TRUE
##  $ stats_logremoval: grouped_df [15 × 13] (S3: grouped_df/tbl_df/tbl/data.frame)
##   ..- attr(*, "groups")= tibble [5 × 5] (S3: tbl_df/tbl/data.frame)
##   .. ..- attr(*, ".drop")= logi TRUE

Thus the user has access to all

5 Visualise results

Finally the results of the QMRA can be visualised for each system component as shown below:

5.1 Inflow 

kwb.qmra::plot_inflow(risk)
Simulated inflow concentrations

Simulated inflow concentrations

5.2 Treatment 

kwb.qmra::plot_reduction(risk)
Simulated reductions in the treatment plant

Simulated reductions in the treatment plant

5.3 Effluent 

kwb.qmra::plot_effluent(risk)
Simulated effluent concentrations

Simulated effluent concentrations

5.4 Exposure 

kwb.qmra::plot_event_dose(risk)
Simulated dose per event

Simulated dose per event 

kwb.qmra::plot_event_volume(risk)
Simulated ingested volume per event

Simulated ingested volume per event

5.5 Health results

5.5.1 Per event 

kwb.qmra::plot_event_infectionProb(risk)
Simulated infection probability per event

Simulated infection probability per event 

kwb.qmra::plot_event_illnessProb(risk)
Simulated illness probability per event

Simulated illness probability per event 

kwb.qmra::plot_event_dalys(risk)
Simulated DALYs per event

Simulated DALYs per event

5.5.2 Total 

kwb.qmra::plot_total_infectionProb(risk)
Simulated total infection probability (for all events)

Simulated total infection probability (for all events) 

kwb.qmra::plot_total_illnessProb(risk)
Simulated total illness probability (for all events

Simulated total illness probability (for all events 

kwb.qmra::plot_total_dalys(risk)
Simulated total DALYs (for all events)

Simulated total DALYs (for all events)

In addtion tables with summary statistics, e.g. for the total risk can be generated easily as shown below:

Total risk (for first repeat of random generation)
repeatID TreatmentSchemeID TreatmentSchemeName PathogenID PathogenName PathogenGroup events inflow_median logreduction_median volume_sum exposure_sum dose_sum infectionProb_sum illnessProb_sum dalys_sum
1 1 Berlin (BF + Slow sand) 3 Campylobacter jejuni and Campylobacter coli Bacteria 730 4956.769 4.040410 1225.871 6148.256 6179 1 1.0000000 0.1295407
1 1 Berlin (BF + Slow sand) 32 Rotavirus Viruses 730 4892.187 3.393638 1225.871 210089.734 210825 1 0.9997758 0.1164240
1 1 Berlin (BF + Slow sand) 36 Giardia duodenalis Protozoa 730 5147.018 1.695504 1225.871 268234.677 268988 1 1.0000000 0.2181268
1 2 Depth & surface filtration 3 Campylobacter jejuni and Campylobacter coli Bacteria 1095 4956.769 2.472764 1838.807 646691.655 646408 1 1.0000000 0.8165979
1 2 Depth & surface filtration 32 Rotavirus Viruses 1095 4892.187 2.591457 1838.807 737878.955 736889 1 1.0000000 0.2380598
1 2 Depth & surface filtration 36 Giardia duodenalis Protozoa 1095 5147.018 1.960714 1838.807 347994.659 347626 1 1.0000000 0.2557319

6 Export results

E.g. Write reports for all configurations in package subfolder extdata/configs/ (here: /Users/runner/work/_temp/Library/kwb.qmra/extdata/configs/):

confDirs <- system.file("extdata/configs/", package = "kwb.qmra")
kwb.qmra::report_workflow(confDirs)