This help page describes the controls and options for plotting monitored data.

Under the simple plotting interface, you are presented with a dropdown menu for Starting and Ending Date.  Data are loaded from the database through the last day of the month chosen in the end date menu.  Selecting an ending date that is before the starting date will result in an error.

For Data Channels, you can select one of the ten data channels in the database, one of the four calculated values, or four sets of common comparison plots.  The four comparison plots are:

1. Generator and Facility Energy
2. Generator and Facility Demand
3. Generator and Facility Gas Use
4. Useful and Unused Heat Recovery

Under the Plot Type there are three types of plots you can choose from:

1. Time Series Plot (Data Channel on the Y-axis, Date on the X-axis)
2. Ambient Temperature (Data Channel on the Y-axis and Ambient Temperature on the X-axis)
3. Hourly Profile Plot (Data Channel on the Y-axis and Hour of Day on the X-axis)

For these plots, all power units (PUs) are plotted as separate lines/symbols on the same plot.

The Additional Options link gives you many more options for plotting data.  The starting and ending date pull down menus are replaced by text fields into which any date can be entered of the form "MM/DD/YYYY" (Month as a number, day as a number and Year as a four-digit number).  Any number of data channels can be selected by holding the <CTRL> key while clicking on selections from the Data Channels menu.  We recommend keeping the selection small as large selections can cause a server timeout during plotting.

The Horizontal Axis menu contains the data channel you wish to plot data against (the independent variable).

The Data Quality pulldown offers three choices:

1. Data Exists - The plot will include data directly from the database, no assurances are provided on usefulness of the data
2. Data Passes Range Checks - The plot will only include data that falls within established levels. 
3. Data Passes Relational Checks - The plot will only include data that not only passes the range check process, but also has passes checks against other corroborating data channels in the database.

For details on the specific Data Quality checks for a given facility, they can be found in the online database notes for that facility under the Facilities tab.  More information on Data Quality can be found in the Data Quality Control section.

The Plot Stacking Method contains options for how the data channels and monitoring units are grouped on the plots.  Three options are displayed:

1. Overlay All Data - One plot that has all data channels from all monitoring units
2. Separate Plots for Each Data Channel - Plots are separated by data channel, multiple monitoring units are displayed per plot
3. Separate Plots for Each Monitored Power Unit - Plots are separated by monitoring unit, multiple data channels per plot

Another option, Multiple Power Units, will appear if more than one monitoring unit is selected.  Two options are displayed:

1. Compare Units will plot each of the monitoring units separately. 
2. Combine Units will combine all the monitoring units into a single data set

The final option is Aggregation Quality, which describes the impact of combining multiple monitoring units on the data.  When data from multiple monitoring units are combined into a single data set, there is the possibility that the data sets do not overlap, or that portions of the data from the individual monitoring units may not pass the Data Quality level desired.  The Aggregation Quality menu contains options that allow you to select the amount of good data displayed on the plot.

The Aggregation Quality menu contains several options:

1. All data - Display all the data from the combined monitoring units, regardless of the number monitoring units represented at each point.
2. Data Valid on at least M of N monitors - Display data for points where M (a number) of N (the total number of monitoring units) pass the desired Data Quality level.
3. Only data that is valid for all N monitors - Display data for points where all monitoring units pass the desired Data Quality level

A higher Aggregation Quality setting typically results in less data on the plot, and in some cases  may be too strict to yield a graph. More information on Aggregation Quality can be found in the Data Quality Control section.

Data Quality Control

There are two methods of data quality control checks that can be applied to the data: Data Quality and Aggregation Quality.

Data Quality filtering applies to the data values as they are loaded from the online database on an hourly basis for an individual monitoring unit.  Aggregation Quality are filtering definitions for how the system combines the hourly data from multiple monitoring units into a single data set.

Data Quality

The Data Quality option is applied to all data when it is loaded from the database.  There are three options for Data Quality from the Additional Options interface: Data Exists, Data Passes Range Checks and Data Passes Relational Checks.

Data Quality Example

The following plot has the Data Quality level set to Data Exists, displaying totally unfiltered data.

The fuel cell at the example site shown has 5 kW nominal output.  One issue which occurred in this data set is the fuel cell output periodically rises above 5 kW.  If this increased power output were actually to occur, a corresponding increase in the fuel cell fuel input would be observed.  It is believed that perturbation in the power data is due to some electrical interference in the data acquisition system.  Based on the nominal fuel cell output, range check for this data was specified to be between 0 and 5.5 kW.

The next plot displays the same data set, with the Data Quality level set to Data Passes Range Checks.  Notice that the points of higher fuel cell output are removed from the plot.

The next example displays the impacts of relational checks on a data set.   The following plot displays the power output for an engine generator, and the corresponding fuel consumption.  The data shown has the Data Quality level set to Data Passes Ranges Checks.  The obvious issue with this data is the power output data remains high while the gas use data drops to zero.

Using a relational check, this data can be filtered from the plot.  The relational check compares   fuel consumption to power output and only returns points where both are valid.  The plot below shows the same range and data as above but has a Data Quality level of Data Passes Relational Checks.  The periods of power production without gas consumption have been removed from the plot.

Aggregation Quality

Aggregation Quality applies to data that is reformed to a larger timestep (daily or monthly) or to combining monitoring units into a single data set. 

Aggregation Quality Example: Single Monitoring Unit

For a single unit reformed into daily or monthly data the Aggregation Quality is a floating point number ranging from 0 to 1 representing the number of data points per day (or month) that pass the desired Data Quality level. 

This example shows how Aggregation Quality is calculated and the impact of Data Quality choices on a sample data set.

Sample data for an example day:

1. Number of records for Data Quality level Data Exists:  24 records
2. Number of records for Data Quality level Data Passes Range Checks:  18 records
3. Number of records for Data Quality level Data Passes Relational Checks:  12 records

The resulting daily Aggregation Quality for each of the three examples would be:

1. Aggregation Quality for Data Exists: 24 records / (24 records/day) = 1.0
2. Aggregation Quality for Data Passes Range Checks: 18 records / (24 records/day) = 0.75
3. Aggregation Quality for Data Passes Relational Checks: 12 records / (24 records/day) = 0.5

The data system allows you to choose the Aggregation Quality desired in integer increments, so the two options displayed for this single monitoring unit data set would be:

1. All data (Aggregation Quality greater than or equal to 0)
2. Data valid on all 1 monitors (Aggregation Quality equal to 1)

Aggregation Quality Example: Multiple Monitoring Unit

For data for multiple monitoring units combined into a single data set, the Aggregation Quality calculation is performed on an hourly, daily, or monthly basis.  The Aggregation Quality is a floating point number ranging from 0 to the total number of monitoring units being combined together.  The Aggregation Quality represents the number of data points per day (or month) that pass the desired Data Quality level for all monitors.

This example shows how Aggregation Quality is calculated for multiple monitors.

Sample data for an example day with three monitoring units being combined together:

1. Monitoring unit #1 has 24 records which pass the desired Data Quality level
2. 2. Monitoring unit #2 has 20 records which pass the desired Data Quality level
3. 3. Monitoring unit #3 has 6 records which pass the desired Data Quality level

The Aggregation Quality is calculated to be: ( (24+20+6) / (3*24) )*3 = 2.08

This indicates that the combined data set for this day represents operation of slightly more than two monitoring units, not all three.  This allows the user to include or exclude periods where data may be bad or missing for a portion of the combined units.

The data system allows you to choose the Aggregation Quality desired in integer increments, so the three options displayed for this single monitoring unit data set would be:

The data system allows you to choose the Aggregation Quality desired in interger increments, so the four options displayed for this single monitoring unit data set would be:

1. All data (Aggregation Quality greater than or equal to 0)
2. Data valid on at least 1 of 3 monitors (Aggregation Quality greater than or equal to 1)
3. Data valid on at least 2 of 3 monitors (Aggregation Quality greater than or equal to 2)
4. Data valid on all 3 monitors (Aggregation Quality equal to 3)

The following plot shows combined generator energy output for two monitoring units with an Aggregation Quality of "Data is valid on 1 of 2 monitors."

In this case, data for one of the monitoring units does not begin until July 1.  Increasing the Aggregation Quality to "Data is valid on all monitors" results in the amount of data in the plot decreasing, as shown in the following plot.  Now the period before July 1 is not shown on the plot and we only see periods where both monitors pass the Aggregation Quality and Data Quality levels.