A couple of points, Speed predictor level crossing controllers are these days standard "off the shelf" hardware and the cost of the controller is quite small compared to the cost of installation. For the this reason one can easily see why such controlers are selected for installation.
Yep! I can understand that they're not a big added expense. But it seems to me that they're still being used where they're not really needed. Presumably they're part of a standard or contract that sees them used in new installations and upgrades in existing ones.
Second point is Speed predictors measure the actual speed of the train and for various reasons this may vary, so one can rely on the crossing being activated in a particular time delay.
Unfortunately, they couldn't be relied on in this instance.
That said, I do see the point of them where there is a variation in train speeds. But between Traralgon and Bairnsdale, the only time you'll see much variation is when a ballast train runs. And the majority of the crossings are not so heavily trafficked that such delays are going to be a problem.
Note, when the hardware is commissioned the trigger speed is checked at all speeds from a crawl to the tracks line speed, the loco going in both directions over the crossing for quite a considerable period of time (well on the Albury line anyway).
Yes, I saw that take place here when predictors were installed. It all worked fine then.
I've been told that part of the problem in Gippsland is that there is actually too wide a contact point between wheel and rail (railhead is worn and too flat on top). Apparently this results in a spread of weight across the rail that means it is insufficient to break through the oxide on the railhead. When the rail grinder eventually arrives and does its thing, the railhead will be slightly rounded, giving a narrower contact point between wheel and rail, increasing pressure to make better contact for the circuits.
Still a bit difficult to believe there isn't enough weight on the rails as it is. But we know similar has occurred in the past, so obviously there's something in it.
In case anyone is interested the Harmon controlers measure the speed of the train by measuring the track circuit impedance at frequnecy in the low audio range (around 400 hz). This frequency can be varied so controlers fitted at closely spaced level crossings can have individual overrlapping trigger distances by the use of the apropriate filters fitted to the track. On the NE line one can see some of the boxes containing the filters around 6000 ft from its level crossing, the filters are also fitted inside the terminating boxes for the insulated joints for the signalling so the audio signal can by pass the insulated joint.
Interesting to know it can be done. Makes me wonder about the crossings here. If so fitted, then they're not working as intended. 1-2 minutes wait at the crossings is not unusual (Princes Hwy, Hobson St and Macalister St, Stratford) when trains stop at the station. Signs indicate down stopping trains not to exceed 40km/h to Princes Hwy (booms will be down before the train actually stops).
The photos below are not directly related, but may be of interest. They show the speed signs for trains either side of Stratford station.
50 and 40 signs for Princes Hwy at Stratford. That's Hobson St in the photo. The highway is out of sight further around the bend.
Yes, 50, 40 and a 65 curve board all within 20 metres. The curve board (still with wooden 35MPH behind it) was recently removed.
And in the other direction
Just before Hobson St, up trains have the 50 sign for Macalister St (not Rd as the sign reads). The just beyond the platform is the 30km/h sign for stopping trains and the works 10km/h sign for the Avon River bridge.
There there's the 65 curve board followed by a 10 curve board and works Commence 10 sign for the bridge.