Halfway through fitting a panel, an electrician is standing at the cable tray holding a cable in his hand. Four cores on it: one 24V, two signals, one spare. The question is simple: which one is which? To find the answer he opens the drawings and works through the pages. One page for the circuit, one for the power distribution, one for the terminals. Ten minutes later he has the answer, pulls the cable, and the same thing repeats for the next one.
On a panel with twenty cables, that is twenty times ten minutes. The bigger the panel, the longer it takes to cable by looking at the drawings, until it is longer than the fitting itself. At the end of the day the cables are pulled, but one question stays with the crew: will I be asking the same question again on the next panel?
Why the cable schedule is a document of its own
The BOM is made for purchasing, the terminal schedule for wiring, the cable schedule for pulling. Three documents, three different readers, none of them at the drawing board at the same time. The person pulling the cable does not sit at the drawing board; he wants one table. Its row is a cable, and its columns are: number, type, core count, from, to, length.
From and to are the two critical columns here. A cable does not run between two circuits, it runs between two sections. You have to think the physical route, not the circuit: from the main panel to the motor control panel, from the control room to the field station. In the drawings that route is scattered over pages; in the table it is one row.
Why a cable list kept in Excel falls apart
In most offices the cable list is put into a spreadsheet in one afternoon after the drawing is done. At the moment of transfer everything is correct; from that moment on the list is an independent copy of the drawing. Every change that comes up while the project goes on makes it into the list "when there is time for it".
The length column is based on estimates. "About 15 metres" means either cutting 15 metres when 16 are needed, or ordering with a margin. Both scenarios are paid for on site. The number column is a sequence typed by hand in a word file: two W7s, a missing W12. As the list grows, each of these errors grows with it, because every row was written by hand.
Core count: the error found when the cable is cut
Let a cable be drawn with four conductors on it, while the cable is declared with three cores. At the drawing board this error is invisible; looking at the page, every conductor seems to be in its place. The error appears when the cable is cut. There is no fourth core. The only two things that can be done at that point: order a new cable, or change the circuit. Both are on site, and both are expensive.
The only reasonable way to catch this error type is to compare the number of assigned conductors with the declared core count. On a project with fifty cables, doing that with the human eye means counting cable by cable, page by page. EupCAD's rule check does this comparison on every run and drops it into the report as "Error: -W3: 4 conductors assigned but the cable is declared with 3 cores". The report carries the cable number; there is nothing left to search for.
Cable numbers: an unnumbered cable is a cable that cannot be pulled
A cable without a number has lost three functions: it cannot be ordered, it cannot be labelled, it cannot be found on site. If a pulled cable has no number, the crew that wants to touch the same tray six months later has no idea what that cable is.
In EupCAD, a conductor being marked as a cable and not yet numbered is a legitimate state while drawing; if the number has not been decided, do not decide it. But that state cannot survive in a released drawing. The rule check reports it per page: "Warning: p.4: 3 connections without a cable number". Which page, how many unnumbered cables, in one line.
When you run the numbering, two properties work for you. First, the run assigns numbers to all marked connections; the manual search-and-label job is over. Second, and more important: cables that already carry a number keep it. Running the numbering a second time does not push W7 to something else; the sequence does not shift, and the labels already out in the field do not become invalid.
Cable type: NYY-J and H05VV-F getting mixed up
The type column is quietly open to errors too. NYY-J for fixed wiring, H05VV-F for flexible connections; the two do not do the same job. If a flexible cable is pulled into a fixed tray, its life is shorter than intended; if a fixed cable is pressed into a flexible machine connection, its safety is lower than intended. If the type is written wrong on the order, the cable tray on site will not accept that cable; the pulling stops for the day.
When the type comes from the drawing, the cable's type is part of the drawing's data. The type is entered once when the cable is added; the same type shows in the list, in the schedule and on the page. Two different types in two places happens when two places are written by hand; it does not happen in two lists produced from one source.
The cable list is read together with the terminal schedule
Every row in the terminal schedule has a cable column: the cable number entering that terminal. Every row in the cable list has from and to. The two documents describe the two ends of a cable. One the terminal side, the other the panel side.
That is why the cabling crew carries two lists. The terminal schedule says which terminal the cable enters; the cable list says where on the other side it lands, which panel, which strip. When the two lists come from the same source, the cable number in one is the same number in the other. A number mismatch happens when two copies were written at two different times; it does not happen in two lists produced from one source.
The Cable Schedule page is part of the drawing
The list is not just a CSV. EupCAD adds a Cable Schedule page to the drawing; the panel package's +KB section sits inside the project as its own page. The drawing that goes to the fitting crew carries the cable schedule with it. There is no such sentence as "there was one more file somewhere".
A new signal that shows up on site
Let the client ask for one more sensor during installation. In the classic flow the electrician pulls one more conductor from an existing cable and does not tell anyone at the drawing desk. Six weeks later, the crew that is about to lay a new cable in the same tray has no idea that conductor exists.
In the EupCAD flow the change is made in the drawing: new signal, new cable data. The cable list is re-downloaded, the Cable Schedule page is updated. The package that goes to the field includes that conductor. The answer to "when did that conductor get in?" is in the package itself.
Length is still an estimate
Let us be honest: do not believe that any drawing tool knows the physical route of a cable. Length is a measurement in the cable schedule too, not the result of a calculation. The difference: when the measurement is written into the list, it is recorded together with the other columns. Core count, type, from, to and length are in the same row; for one of them to go stale, you would have to do a separate job for the others, and that is mostly not done.