At the end of the Analysis, InfoDrainage evaluates how "hea= lthy" the simulation was (for each Storm and each Phase). To do this, InfoD= rainage relies on a number of indicators calculated by the engine.

Flow Cont=
inuity |
Comparison of the= total incoming volume (including initial volume) and the total outgoing vo= lume (including the final volume) at the end of the simulation. A discrepan= cy between the 2 volumes indicate that some volume has been either lost or = created during the simulation. A significant value of Flow Continuity error= implies a poor confidence in the flow results. Because the pollutant trans= port is dependent on the flow, a large Flow Continuity error normally resul= ts in a large Quality Continuity error. |
---|---|

Quality Continuity |
Comparison of the total incoming pollutant mass = (including initial mass) and the total outgoing pollutant mass (including t= he final mass) at the end of the simulation. A discrepancy between the 2 ma= sses indicate that some pollutant mass has been either lost or created duri= ng the simulation. A significant value of Quality Continuity error implies = a poor confidence in the concentration results. |

Convergence |
At each computational time-step, the engine uses= an iterative procedure to calculate the updated hydraulic state. After eac= h iteration, the engine checks for the "convergence" of the water levels wi= thin the system to determine if the new hydraulic state is acceptable or no= t. There is a maximum number of iterations allowed to reach convergence (8 = iterations by default). If the convergence has not been reached after 8 ite= rations, the engine keeps the hydraulic state from the last iteration and m= oves on to the next time-step. However, in this situation there is a high c= hance that this hydraulic state is not physically valid. |

Depending on which indicator is flagged by InfoDrainage as inappropriate= , the possible ways of improving the simulation health are described below.=

Typically, the time-step is too large, or an outlet is ignored by the ti= me-step condition.

Innovyze recommends that you consider selecting a shorter time step on the **Analysis Criteria**. =
This reduces the upper bound of the time-step. The engine still has the abi=
lity to adjust the time-step within the prescribed range according to the f=
low conditions.

Time-step range |
Time-step lower bound (s) |
Time-step upper bound (s) |
---|---|---|

Default |
0.5 |
15 |

Reduced |
0.5 |
5 |

Shortest |
0.5 |
1 |

If both the Flow Continuity and Quality Continuity error are high, then = the Flow Continuity should be addressed first (see section above). In most = scenarios that will also improve the Quality Continuity.

If the Flow Continuity error is low and the Quality Continuity is high, = it is recommended to contact the Innovyze Support team for further investig= ation.

An object or structure within the project is making the convergence diff= icult. This should be addressed first, only then consider reducing the time= -step range

The following checks or actions should be considered.

- Ensure that inflows into the system are gradually-varying, i.e. avoid a= jump from zero inflow to 100 L/s over a short period of time for example.<= /li>
- Simple Junction nodes should be used in places where the hydraulic cond= itions are fairly simple. Consider replacing a Simple Junction by a Manhole= where more than 2 connections are meeting, or where a node becomes highly = surcharged during the simulation.
- Outlets need to be treated with caution. The adaptive time-step conditi=
on currently ignores outlets (weirs, orifices, gates, etc). If an outlet is=
behaving badly, it is necessary to reduce the time-step range.

Further Help

In the case that none of the measures described in this page improve the= simulation health, please contact the Innovyze Support team to get further= help.