Skip to content

Time Series with NaNs (empty data points)

Navigation:

Meaning of NaNs

NaN is the abbreviation for “Not a Number”. There are no valid numerical values at the times to which NaN is assigned. NaNs indicate empty data points of a time series.

NaN stands for a numeric data type value that corresponds to an undefined or unrepresentable value.

Display of NaNs

In the plot of the time series in ETA and in the plot preview in TOP-Energy NaNs are displayed with grey bars (see following slider).

Plot and value table with NaNs

TOP-Energy plot preview with NaNs

In the value tables (data grid) of the time series, NaNs are first displayed with the term Empty. After double-clicking the Empty entry, the cell display changes to nan (see following figure).

If time series already contain values that cannot be interpreted during import into the ETA, these are shown there as NaNs.

NaNs in Input Time Series

Empty values (NaNs) are present in input time series if, for example, measured values are missing at certain times because measuring instruments have failed. The linked article describes how NaNs are handled in the input time series during import.

Empty input values result in empty output values for these points in time. Depending on the application, it can be decided whether the input values are to be corrected in advance (deleted or interpolated) or whether the failure of measuring instruments is to be mapped realistically with empty values.

In the simulator window, error message 8271 reports the empty data points of an input time series (see following figure).

NaNs in Output Time Series

There can be several reasons for the occurrence of NaNs in the output time series. The two main causes for the emergence of NaNs in output time series are lack of solutions and the shutdown of the plant.

Optimization Problem Not Solved for this Point in Time

NaNs are output for a point in time when the optimization problem cannot be solved for this point in time. This occurs, for example, when missing input data prevent the calculation, when the capacities are not sufficient to cover the demand, or when operational boundary conditions restrict the mode of operation to such an extent that operation is no longer possible. In these cases, the error message 8355 appears in the simulator window, and the simulation closes with the following message (see following figure).

If the NaNs occur at powers or energies, it is to be assumed in principle that the optimization problem has not been solved. You should then search for the cause in order to eliminate the error (if necessary, using the Balance Relaxation components).

Plant Not in Operation

Intensive material values (e.g., temperatures) are not calculated if the system, e.g., a hot water boiler, is not in operation. A switched-off plant has no air and exhaust gas mass flows, for which temperatures could be calculated. In these cases, it makes sense not to estimate the temperatures because this would result in incorrect annual average values.

It is easy to check whether the empty spaces (NaNs) result from the standstill of the plant. For this purpose, the plot of the intensive quantity, e.g., the exhaust gas temperature, is compared with the plot of generation capacity or the operating hours of the same component (see following figure). If the grey bars only appear at times when the power is zero, i.e., the component is not in operation, this is correct and poses no problem.

Back To Top