Question
How to Reduce Run Time of MALT and MALZ?
Answer
There are several ways to significantly shorten the run time, even for resource-hungry, complex, MALT and MALZ models. The following factors affect run time:
1. The number of conductor segments modeled. Modeling of dense grids at switches, rebar in foundations, or short conductors to equipment can increase run time needlessly. Typically, modeling a small subset of the conductors in dense grids and foundations demonstrates satisfactory safety: modeling all of the conductors is overkill (if an area is safe with a subset of the grounding conductors present, then it most certainly is with more conductors). This approximation has little effect on the rest of the grounding grid.
2. When a large number of conductor segments are modeled, much of the computer's RAM is used. If the computer has insufficient available RAM, either because the computer has little to begin with or because other memory-hungry applications are running, then swapping to the hard disk occurs and the run time skyrockets. If swapping is occurring, shutting off certain applications during MALT and MALZ runs involving large numbers of conductors may reduce the run time tremendously.
3. The soil structure has a great impact on the run time, because it determines how many series terms must be computed for each and every earth and conductor potential. High contrast ratios between layer resistivities increase run time, as do the presence of very thin layers (compared to the thicknesses of other layers) and a large number of layers. By inspection, it may be possible to eliminate thin layers which have little real effect on the performance of the grounding system, but do have a great effect upon run time.
4. When all conductors are horizontal and at the same depth, all MALT and MALZ potential computations employ acceleration routines that can reduce run time by an order of magnitude. In large grounding systems, short ground rods (e.g., 10 ft) typically have little effect on the performance of the grounding system and can usually be omitted, at least during the initial design iterations. Similarly, small differences in burial depths of conductors, say in the switchyard versus the power plant, may result in negligible differences in grid performance, but very noticeable effects on run time.
5. The default potential computation accuracy setting is 1%, set in the Computations/Advanced screen. Loosening the accuracy to 5% (which must be entered in decimal form as 0.05) can speed up the runs considerably as well, and should be done during the initial design iterations.
6. The number of earth surface potential computation points also affects run time: computing potentials at 100,000 points takes 10 times as long as computing potentials at 10,000 points. During the initial design iterations, the designer can afford to space points relatively far apart, increasing the density in areas of particular concern or during the final design verification run.
No Related Articles Available.
No Attachments Available.
No Related Links Available.
No user comments available for this article.