++systems
kanal++
contact
imprint
english version

Sewerage network calculation


Historical outline of the methods in sewer network calculation


  • Time coefficient method:
    Originally this method was just about suitable dimensioning of pipe diameters. It is based on a rainfall intensity curve, of a provided yearly frequency or return period regarding rainfall intensities in ascending duration. In this regard, at any point of the sewer network the relevant intensity is exactly the one, the duration of which corresponds to the flow time. The results are to be considered independently from the depth of the sewer network. Therefore this method is absolutely not suitable for flooding verification or dimensioning of overflow basins.

  • Constant runoff coefficient or cumulative frequency polygon method::
    Basin dimensioning is possible, though no overflow verification.

  • Variable runoff coefficient (Pecher):
    Capable of basin dimensioning.

  • Surface runoff model:
    Basin dimensioning and limited flooding verification are possible. As the model is temporally asymmetrical, individual concepts are useless and not economically reasonable, as backwater conditions are not reproduced realistically. Carrying out suitable calibrations, method and data inaccuracies are tried to be counteracted against. In this way calibration is intended to avoid exponentiation of errors in the results.

  • Temporally symmetrical sewer network calculation:
    This procedure distinguishes itself from temporally asymmetrical calculation by the sewer network being calculated as a whole at any point in time, instead of being calculated chronologically from the top downwards, or the other way round. If the network calculation is aborted before having been completed, the entire network has been calculated completely and evenly till the point in time of the simulation abort. As regards temporally asymmetrical sewer network calculation, also known as hydrological methods, only the upper part of the network is calculated entirely, though covering the whole simulation period. If the network load at any point in time was represented in a movie, with a temporally symmetrical method it would be impossible to recognize whether it runs forward or backwards. In case of an asymmetrical procedure this would of course be immediately visible, as rewinding, the downward parts of the network would fill first and the uppermost ones in the end. It is self-evident that temporally symmetrical methods are distinguished by higher accuracy, as all network conditions can be surveyed synchronously. The updated network condition results with very fine temporal resolution from the previous condition by elementary mathematical solution of an initial or boundary value problem, formulated observing the volume, energy and impulse balances. Unlike asymmetrical procedures, these temporally symmetrical methods are very calculation time demanding. For this reason, in the last years, new ground has been broken in this regard, introducing state-of-the-art multi-core processors (symmetrical multiprocessing), as well as distributed calculations on several computers (massive parallel calculation). Tandler.com has been pioneer and precursor in this matter by the development of the complex parallel step method within the sewer network calculation system KANAL+ hydraulics. With this solution, calculation times could be reduced drastically, so that it is now possible to carry out serial and long term simulations with individual area data, taking into account and verifying the requirements of water protection, as well as of flooding safety at the same time. A lot of practical studies have shown that by now, calibrations are just necessary to correct excessively rough or inaccurate data within the model, by means of precise measurements. These expenses for calibration will be reduced considerably, if increasingly precise and individualized data is included into the models (uneven rainfall coverage, soil composition, surface characteristics), as long time experience has shown. For long term prediction of the system performance, data individualization is not absolutely necessary, in order to obtain an unbiased and safe estimation concerning long term load and flooding probabilities. Raising the security parameters, this is the case, for instance, if the catchment area is covered by one large-scale area rainfall, instead of being provided with a network of precipitation measurement stations. The securities reached in this way are distributed more evenly, however, than by trying to obtain them by means of a method afflicted with deterministic errors – as caused by temporal asymmetry. In the end, this is a cost factor, too.

Progress in sewer network calculation as delineated here is mirrored almost entirely in the software provided by us.

KANAL++