Analyzing WRIC Data: Zero Tests and Methanol Burns • wrictools

library(wrictools)

Introduction

This vignette shows how to analyze Zero Tests and Methanol Burn experiments using analyse_zero_test() and analyse_methanol_burn().

Zero Test Analysis

A zero test checks the baseline VO2 and VCO2 of the WRIC system, the mean should be 0 and the standard deviation low and both should be constant over time. Use analyse_zero_test() with a data file to get plots and statistics.

# Example file included with the package
filepath <- system.file("extdata", "data.txt", package = "wrictools")

# Analyse zero test
zero_stats <- analyse_zero_test(filepath)


print(zero_stats)
#> $room1
#>   variable     mean        sd       min      max       slope
#> 1      VO2 356.7205 100.54488 -29.90960 786.6187 0.002766024
#> 2     VCO2 298.4697  87.95359 -45.17043 674.9738 0.002499467
#> 
#> $room2
#>   variable     mean        sd       min      max       slope
#> 1      VO2 332.0050 122.09571 -500.9322 708.2423 0.001944661
#> 2     VCO2 267.9337  99.32789 -413.3808 565.4705 0.001476097

As you can see the example data provided with this package does not look like a zero test, which you can also see when looking at the statistics. This vignette is intended as a quick example on how to use the package. If you change the file to your own zero test you can quickly assess whether the mean is 0, the standard deviation or whether there is an offset.

Methanol Burn Analysis

Methanol burns validate energy measurements in the WRIC. Use analyse_methanol_burn() with a WRIC data file and a methanol measurement file. The measurement file needs two columns named datetimeand methanol. Datetime by default assume this format “2023-11-13 22:00”, but you can specify your own dateformat and give it as a parameter to the function.

# Example WRIC file
data_txt <- system.file("extdata", "data.txt", package = "wrictools")

# Made up methanol measurements to show functionality
methanol_df <- data.frame(
  datetime = as.POSIXct(c(
    "2023-11-13 22:00", "2023-11-13 23:30", "2023-11-14 03:00", "2023-11-14 06:00", "2023-11-14 08:00"
  ), format = "%Y-%m-%d %H:%M"),
  methanol = c(4842.9, 4785.7, 4758.6, 4675.1, 4629.2)
)
methanol_file <- tempfile(fileext = ".csv")
write.csv(methanol_df, methanol_file, row.names = FALSE)

# Analyse methanol burn
methanol_results <- analyse_methanol_burn(filepath = data_txt, methanolfilepath = methanol_file)

print(methanol_results)
#> $per_interval
#> # A tibble: 4 × 14
#>   t1                  t2                  delta_methanol_g delta_CO2_L
#>   <dttm>              <dttm>                         <dbl>       <dbl>
#> 1 2023-11-13 22:00:00 2023-11-13 23:30:00             57.2        40.0
#> 2 2023-11-13 23:30:00 2023-11-14 03:00:00             27.1        19.0
#> 3 2023-11-14 03:00:00 2023-11-14 06:00:00             83.5        58.4
#> 4 2023-11-14 06:00:00 2023-11-14 08:00:00             45.9        32.1
#> # ℹ 10 more variables: delta_O2_L <dbl>, delta_time_min <dbl>,
#> #   CO2_ml_min <dbl>, O2_ml_min <dbl>, methanol_g_min <dbl>,
#> #   VO2_measured <dbl>, VCO2_measured <dbl>, VO2_dev <dbl>, VCO2_dev <dbl>,
#> #   RER <dbl>
#> 
#> $overall
#> # A tibble: 1 × 7
#>   VO2_avg_meas VCO2_avg_meas O2_avg_calc CO2_avg_calc VO2_dev_avg VCO2_dev_avg
#>          <dbl>         <dbl>       <dbl>        <dbl>       <dbl>        <dbl>
#> 1         349.          292.        423.         282.       0.153        0.442
#> # ℹ 1 more variable: RER_avg <dbl>
#> 
#> $plots
#> $plots$O2

#> 
#> $plots$CO2

#> 
#> $plots$RER

#> 
#> $plots$burn_rate

Again these results are non-sense because the wric-data we’re analyzing is not a methanol burn and the methanol measurements are fictional as well. But you can easily see the information the function returns.

We can see the different measurements and most importantly the deviation between calculated and measured VO2 and VCO2 for each interval, meaning each time period between two measurements in the methanol measurement file. This deviation should ideally be 0. If it is not, check whether this might be at the beginning or end of the file where a door or hatch might have been opened. There is an additional table showing the same measurements for the entire period - if you use these values make sure to cut of any time in the beginning or end that might show noise from the door opening. You can easily do that by specifying start and end parameters.

Additionally, 4 plots are created showing predicted vs measured O2 and CO2 over time. It is important to note that the value from one interval is plotted at the end point of that interval. The other two plots show RER over time and the methanol burn rate.