The world's longest answer...
“The three most important things in civil engineering are drainage, drainage and drainage.”
- Bryan Duff, P.E.
As to the question regarding tie placement on ballasted deck bridges with a skew built in, in the few cases I’ve seen the angle of skew is not very great and only a single tie on either side of the gap between bridge and terra firma is askew. Generally, the angle of the askew ties matches the angle of the bridge. It wouldn’t take much of an angle, perhaps 15° or so, before placing ties across the gap would become the only option.
As far as the layout of your abutments is concerned, I think the volume of water the stream sees at peak periods plays a key role in the position of the abutments and the shape they take. The direction of flow and any redirection that occurs would change the shape of the abutments as much as the surrounding terrain, slope, soil types, etc.
From your description of the stream, it sounds like some change in direction occurs as the stream approaches the bridge. The outer portion of the stream's curve would need to be protected against erosion, since there is a greater probability erosion will occur there. Likewise, the inner portion of the curve should be excavated more steeply and the abutments reinforced there as well, perhaps with a concrete apron, to permit flow and prevent alluvial action.
With respect to the type of abutments to be constructed, I think a reconstruction effort on the scale you're talking about, i.e., a vertical re-adjustment of the alignment, would necessitate totally new abutments. It's possible new abutments could be constructed on top of the older ones, but based on your description of the traffic the rail line saw in the past, I doubt it would have been built to support the existing traffic seen on the line at the time of the construction project, let alone the weight of a new alignment built on top of the old one. So, I surmise that it would be cheaper to construct a single twin track abutment and embankment than two separate ones, despite the tracks being separated by 22 feet. Moving earth to create the subgrade for an elevated earthen right-of-way such as this, gaining compaction, etc., is quite an undertaking and could be simplified by treating both sets of track as a single unit for the purpose of this phase of construction.
However, all bets are off if traffic has to be allowed for during the course of construction. In that case, a temporary third track might be constructed outside the limits of the two existing tracks to allow for reconstruction of one of the tracks at the new vertical alignment. Upon completion of the new vertical alignment, the second track would be constructed. Its completion would allow for the removal of the temporary third track.
Nonetheless, a single monolithic abutment, whether constructed at once in a single pour or in a modular fashion, would be the best choice for abutments with adjacent tracks being as close together as 22 feet. A drainage swale would need to be provided between the tracks outflowing to the stream. Once the runoff reaches the abutments, it would most likely be removed by sheet flowing the water over the top of the abutment and would likely include a concrete apron to facilitate this.
Now, if your layout depicted the tracks before the vertical alignment was changed, there is definitely reason to believe that two abutments, most likely of a different construction style, if not technique, would exist in a fashion similar to what you have. I believe they are undersized, in terms of the thickness of the concrete, though. I would expect them to be no less than 12 inches in width at the top for the size abutments you have in place. The one thing that bothers me about the abutments as you have them currently is the crook between the two on the right, where the wing walls converge. That creates an area on the outside of the curve of the stream where water will swirl and rapidly erode the subgrade. The outside of the curve of a stream should be free of any impediment to directionalizing the flow of water downstream, although it’s not uncommon to find drainage structures such as gabions placed along the stream bottom to slow the speed of flow. The point here is that creating a turbulent situation which favors erosion is bad. Creating an impediment to slow the flow is good. So, even though the scene as depicted in the photo is not intended to be the earlier version of the alignment (i.e., in the floodplain), it could easily pass for such a design. For the modern version of the abutment much of the philosophy stated here was considered in my design recommendation of a single monolithic abutment vs. two independent abutments.
You mentioned the vertical alignment change affected how wide the footprint of the embankment is. You are correct in anticipating this and correct in forseeing one of the solutions to mitigate the problem could bea concrete retaining wall. There are other solutions that are cheaper than a concrete retaining wall, but sometimes what’s most effective isn’t cheap. You can do it right and pay for it once or do it wrong and pay for it twice. So, with that in mind, here’s a couple guidelines to designing prototypical drainage side slopes.
Based on the surveying I did for ATSF, there are three distinct side slope situations you can have: shallower than 1:1 (commonly 2:1 on high embankments), 1:1 slope, steeper than 1:1 slope. An earthen embankment such as what you've described can be constructed without reinforcement with a 2:1 or shallower side slope. Where water would run roughly parallel to the embankment, the portion of the embankment that will get wet would have riprap placed up to the stream bank to slow erosion. Other than that, no soil stabilization would be necessary other than vegetation.
If the catch point (where the design side slope intersects the existing ground) prevents a 2:1 side slope from being used, up to a 1:1 slope could be used with reinforcement, such as steel or concrete pilings,
gabions, or a concrete apron. 1:1 is the magic number, so to speak, since the load at the top of the embankment is borne by the earth underneath it diminishing downward at a 45° angle on each side. The 1:1 slope means that the outer limits of the embankment is also where the load bearing portion of the embankment ends. A 1:1 slope isn't unstable, but since it's at the limit of design, it should be reinforced.
Any slope steeper than 1:1 requires some serious reinforcement, since the load must be borne in an area smaller than is necessary. Engineered retaining walls come into play here and they are expensive to construct. They must have a rather large, deep footing, and when they serve as a bank of a stream in addition to retaining earth and bearing a load, they must be impervious to water. If that's the case, they are generally monolithic slabs poured in place. If the catch point impacts the stream but doesn't necessarily cause the retaining wall to act as the bank of the stream, I think it would be possible to construct the wall out of an interlocking concrete material, such as what is used on modern interstate bridges. Walls can even be constructed out of gabions, but I can't think of a case where the gabion wall is load bearing (i.e., an engineered wall).
So, what do I think is the best solution for your bridge dilemma? Use your existing deck bridge and get an identical one for the other track. Construct a single abutment on either side of the stream, including aprons at the tops of the abutments to provide for between track drainage, and funnelling aprons on the upstream side of the embankment (small and low on the inside of the curve, large and high on the outside of the curve). Any portion of the track that runs parallel to the stream should have no less than a 2:1 slope from the top of the roadbed shoulder to the natural ground catch point if it is unstabilized. If it's less than 1:1 but greater than 2:1, consider gabions, riprap, steel or concrete pilings or any combination thereof. If the slope is greater than 1:1, construct concrete retaining walls. If the retaining wall stays wet, make it a paved wall. If it stays dry, consider using a modular style wall.
I hope you found this post informative and you're able to put some of this knowledge to work for you designing your bridge solutions.