If it's VSD driven, the problem could well be protective conductor currents ... although radiated emissions may also be an issue.

The SY braiding and trunking will only be effective barriers to radiated emissions, if they are properly earthed, and provide complete 360-degree coverage (including with the SY, use of proper glands). Poorly fastened lids, and corrosion, of the trunking, may also affect things. Physical separation, of say 1m, may help.

If the problem is protective conductor currents, then additional earthing and bonding is a good solution - see BS EN 50310 - but over such long distances a good solution may be costly to implement.

Another possibility is conducted EMI through mains - this can be more problematic, and might require dedicated noise filters to address.

Thank you.

The sheath of the SY (for what it's worth), is connected to earth at the supply end (the outlet to the machine is a plastic 5 pin industrial connector).

Local earth rod near the machine?  I am dubious that the earth at the DB is all that great, due the distance.

I think I will start by moving the cable 1m away from the rest of the cabling.

The fluorescent lights in the offices on the mezzanine nearby flicker when the machine pulls a high load.

Before un-routeing a load  of cable, if you can, see how much the voltage on the wiring to the lights is wobbling about - it may be good old fashioned voltage drop, and 'conducted emissions' rather than a radiated EM .
If the motor draws 30A phase when up to speed it may be trying, and failing, to get a few times that during start up and changes of speed . If so, Mr Ohms law may be enough to explain a lot of the rest. You will need a voltage logger or scope (can be hired) or at least a meter with a very fast response or max/min hold.

Is it easier to lay in a new main for the lights and delicates , rather than the motor ?
Mike

Rewiring the lights would be a nightmare. 

A new feed to the machine would be easier (well... it ain't easy, but hey ho).

Might see what I can assess with some supply monitoring.  The machine runs fine, but working in the offices when it's running makes people turn off the lights.

Or change the fluorescent lights to something less sensitive to voltage variations - e.g. LED ?

    - Andy.

I agree that voltage drop may well the reason for the observed lamp flicker. Also agree that changing the fittings to LED alternatives is well worth considering. Try changing one fitting initially.  LED lights will probably pay for themselves by energy saving.

OK.  I have done some investigating.

Lighting is wired in some pretty shoddy 1.5mm and fused at 20A.  I would assume that is a throwback to the days when they had sodium lights?  Don't think I have ever seen a lighting circuit protected by 20A.

The company had someone come in and check all the wiring and said all was fine.  But it isn't.  DB lists are wrong for starters.

Question.  As a qualified, 55 yrs old lifelong sparky but who is no longer registered with the Niceic (because I no longer install - quit 2 years back), am I OK trying to sort out this issue?

I was told by this highly suspect electrical company that as it's a industrial site, it doesn't need to be signed off.  Not sure that is correct.

But, I do know that we don't want them back, and I would do the job properly.  Just wondered what the regs were these days, seeing as they change every 15 minutes.

Don't think I have ever seen a lighting circuit protected by 20A

Not unknown - I've seen a few 1980s office blocks with 20A lighting circuits with 20A grid switches and fluorescents with fused terminal blocks - usually in 2.5mm² though.

I was told by this highly suspect electrical company that as it's a industrial site, it doesn't need to be signed off.  Not sure that is correct.

Partly - all the troublesome Part P of the building regulations and the corresponding registered competent persons scheme and notification stuff is only for domestics. For everything else it's vanilla BS 7671 plus however the local duty holder decides how to demonstrate how they're meeting the Health & Safety at Work legislation and the Electricity at Work regulations (and associated legislation, and possibly insurance company requirements too). A sensible minimum is likely to be BS 7671 certificates signed by whichever competent person did the work. A lot of companies like NICEIC or similar registration as a means to demonstrate that those doing the work are competent (or if it turns out they're not, there's someone else to blame!)

   - Andy.

Assuming you are happy to do the testing, mainly verify insulation, earthing etc and sign on the minor works cert that is is OK, that in legally (maybe actually more than) enough. No need for anyone to demand to see your certificates or to pay club memberships.
But the buildings insurance or the local management may care, and impose house rules, anti-discrimination rules do not allow you to be barred on your age but an arbitrary 'NICIEC only'  requirement may be in place. You will have to ask.

It does not have to be your day  job... As an example I neither belong to a part P scheme nor the IET, but I have designed and installed a fair amount(!). I just have to jump hoops for domestic local authority notifiable, so tend to avoid that except for family and friends. Equally, (as I have the required knowledge and the self confidence, and access to the test kit), I have done plenty of new circuits and things for scouts HQs and similar situations (as a charitable organization I strongly support) where there is a local ownership of the buildings and some common sense can be applied, and as required I then fill out my own certs for the duty holder and exec committee to refer to. For me the day job tends to involve larger project teams, and I may draw things in the plans, but someone else can do it then.
Mike.

A 20A lighting circuit sounds a bit on the high side,  though 16A driving ES lamp fittings used to be quite common in the days of big bulbs .  (1960s 250W filament lamp anyone ?)
Would be typical if the light fittings plugged in with 13A plugs, but then it is a socket circuit with switches, not a lighting cct. In any case not against the rules, just odd.

But, if it is 1.5mm it may be a bit undersized depending on cable grouping factors and so on,  (or if you prefer, a bit over-fused), but in practice most unlikley to be an issue - and it may have been done to avoid a need for C or D type breakers rated at a lower current.

Mike