If you’re an old hand at PLS-CADD™, terms like L1 through L4 are second nature and need no explanation. If you’re relatively new to the software, my advice is different: don’t use them, at least not yet. Ignore the fact that L1 through L4 – or their structural cousins, M1 through M4 – exist at all.
Don’t call them by the abbreviations, which only invite misunderstanding. Call them by the purpose of the analysis you’re actually performing instead. Do that for a while, and you’ll find those L’s and M’s stop being an arcane pursuit and start being convenient shorthand for something you already understand.
Structures Have Their Methods, Conductors Have Their Levels
PLS-CADD™ uses a similar numbered shorthand for how it models the wires themselves – the “conductor system” – as it does for structures. A conductor system covers everything that defines how a conductor behaves: the structures it’s strung between, the insulators, the conductors themselves, any marker balls, and the loads applied to them. Where structures have Methods 1 through 4, conductor systems have Levels 1 through 4 – L1 through L4 – and it’s those four levels of sag-tension modelling sophistication I want to walk through here.
Before we get into them, a few terms are worth defining properly.
Span Length is the horizontal (2D) distance between the centre of one structure and the centre of the next. It can differ from the circuit span length and from the “inclined span length.”
Inclined Span Length is the true (3D) distance between the centre of one structure and the centre of the next.
Tension Section is the distance between dead ends – it should match the sum of the span lengths between dead-end structures.
Ruling or Equivalent Span is a weighted average, calculated as:
RS = √( (S₁⁴/C₁ + S₂⁴/C₂ + … + Sₙ⁴/Cₙ) / (C₁ + C₂ + … + Cₙ) )
where Sᵢ is the horizontal (2D) span length and Cᵢ is the true (3D) inclined span length.
Level 1: The Ruling Span Approximation
Open PLS-CADD’s own demo.xyz example file and go to Sections → Table, and you’ll see each section’s display condition listed as Initial or Max Sag RS – RS standing for Ruling Span. That tells you the sag-tension behaviour across that section is being defined by the ruling span approximation: Level 1 analysis.
If you don’t have the SAPS option, every engineering study and analysis you run will be in Ruling Span mode by default – Level 1 in SAPS terminology. Under this approach, tension stays the same throughout the whole tension section, but there’s no interaction between tension sections. Dead-end attachment points are treated as fixed and won’t move, so none of the structures are given any flexibility. Because tension stays equal throughout the section, there’s no differential tension across a suspension insulator, and therefore no insulator swing.
Power Line Systems themselves consider the Ruling Span approach suitable for the vast majority of engineering applications. I have to disagree, and say that Level 2 is more appropriate for most of the work I see – unless you’re working somewhere unusually flat and uniform, where span lengths and terrain barely vary from one structure to the next. Oklahoma is the textbook example. In that kind of territory, the ruling span assumption holds up fine, because there’s little real difference between an “equivalent” span and the actual ones either side of it. The more your spans and terrain vary, the less that assumption holds.
Level 2: Finite Element Modelling Without Interaction Between Wires
In Level 2, all supports are treated as infinitely rigid, unless you deliberately add fictitious springs – which in practice means stepping up to a full Method 4 structural model: PLS-CADD’s most detailed way of defining a structure’s strength, component by component, rather than treating the whole structure as one rigid unit. The method also ignores interaction between wires, whether that’s between adjacent tension sections or between other phases on the same section. What it does allow is differential tension within a tension section: a tension section is only ever perfectly equal at one condition, usually the one it was installed at, and any change in temperature or creep afterwards will leave different amounts of slack in each span. That slack changes the tension in that span, and it’s most noticeable when a suspension insulator starts to lean and the sags visibly vary.
This matters most where you have a long span sitting next to a short one. Having worked through thousands of miles of LiDAR data – first approximating with Ruling Span, then using finite element modelling to match what the conductor was actually doing once installed – I can say that very few real lines are accurately represented by the Ruling Span approach alone.
Level 3: Finite Element Modelling With Interaction Between Wires
Level 3 gives results similar to Level 4, but computes faster, because it uses a simplified stiffness matrix built when each structure’s full Method 4 model is analysed – avoiding the iterative full structural analysis that Level 4 needs to converge.
My own rule of thumb: if you’re working with slender structures likely to deflect, with guyed structures, or trying to mimic the effect of stringing operations, use Level 3. For most other situations, Level 2 is sufficient. A good practical compromise is to analyse all PLS-POLE structures at Level 3 while treating TOWER models as Level 2 – and if you want to go further still, PLS-CADD™ lets you limit the enhanced Level 3/4 analysis to only the structures within a chosen number of spans of the one you’re actually computing loads for, rather than running it across an entire line.
Level 4: Full Interaction of All Wires and Structures
Level 4 is the most advanced modelling level available: a full structural analysis of the entire tension section, including detailed models of every supporting structure and every cable. Is it the right choice? Usually not. Because it’s computationally intensive and rarely justified, Level 4 should be reserved for special cases where you genuinely need an accurate picture of the interaction between structures and wires. Most engineers will rarely, if ever, need it.
Conclusion
Of the four levels, Level 1 is the one to reach for only where terrain and span lengths are genuinely uniform; everywhere else, it risks hiding real differential tension and insulator swing that a Level 2 analysis would catch. Level 2 is the right default for most day-to-day work. Level 3 earns its extra computation time on slender or guyed structures, or where you’re trying to mimic real stringing behaviour. Level 4 is there for the rare case that genuinely needs it, not as a default setting. As with PLS-CADD’s structure methods, the right choice here comes down to matching the level of sophistication to what your structures and terrain actually demand – not to reaching for the most advanced option out of caution.
As Director of Engineering and heading up research and development at NM Group, Paul Richardson is a renowned power industry expert and long-time proponent of the use of 3D modelling in improving reliability and utilisation, while minimising cost. He is a chartered civil engineer (BEng Hons, CEng, MICE), and retains a wealth of experience, including more than 30 years of power line engineering.