Are you using clarified or cloudy concentrates?
If your recipe includes the latter, then part of the precipitate will almost certainly just be settling of the natural suspended fruit pulp that gives cloudy juices their turbid appearance.
Assuming you're working with clear concentrates, these will typically be <5NTU at single strength but may be higher for some fruits and from some processors, so there will still potentially be some suspended insoluble matter in there that can settle after bottling.
This link is focussed on apple, but gives a reasonable overview of common mechanisms of haze formation in juices of most types: https://www.ncbi.nlm.../pubmed/9067089 (It's only the abstract, but gives you an idea of some topics to research in more detail).
Even if you're starting with exceptionally clear concentrates, you can find that by combining two juices you change various chemical potentials within the overall solution and this can lead to formation of various solids / semi-solids that cause a haze or settle into a precipitate. In fact, even with just one juice concentrate this effect can be seen when bottling, as the composition of the water used can have quite a significant impact.
You may find this only happens irregularly, due to variation in the water source and/or the natural variations in the fruit used to make the concentrates.
If you're using significant quantities of grape concentrates in any products then this can do peculiar things all of its own, as the naturally high tartaric acid and glucose content mean it can be prone to crystallisation of glucose and/or precipitation of potassium hydrogen tartrate, particularly if exposed to significant temperature changes.
We work with a lot of juices and seem to be seeing haze problems more frequently at the moment - I actually think that people moving away from using gelatine as a fining agent may produce concentrates that don't always perform as well, as these days we find that many processors simply throw the juice through a fine crossflow filter or similar, which will "clarify" it in the sense that it removes the bits that make it cloudy, but doesn't stabilise it in the same way as a proper fining process would. This seems to be combined with an attitude of "we're using the public mains water supply so of course it's fine", and then people are surprised that things aren't behaving as they expected...
Without knowing more details about your particular issue it's difficult to suggest specific solutions, but you may want to look at the following as some potential considerations:
- Enzymatic treatment - pectin can be a common source of issues in some types of fruit, so pectinase can be used to remove this. Whether it's necessary will depend on the nature of your ingredients and their initial processing, so this should be discussed with your suppliers in terms of whether they do this already, and if they test for it. You can do a simple check with your products to see if it's potentially an issue - either take a finished product, or a sample of the concentrates that you've reconstituted to single strength, and add 10ml of the sample to 10ml of ethanol in a test tube, mixing gently (don't shake!), then leave to stand for a few minutes. If you see lumps forming, or a thick "skin" on the surface, then there is a reasonable amount of pectin present and this could be part of the problem you're having.
- A filtration stage may remove some of the haze, but on its own it probably won't cure the problem.
- You may need to combine a fining regime with the filtration step, as this will potentially be able to remove both the haze and the source of the problem, giving you a finished juice that is much more stable throughout shelf life. There are several potential options for fining agents - gelatine and isinglass are both effective in many cases, but for some markets these are not very popular as they result in a juice that isn't suitable for vegans, and of course some sources of gelatine can also be problematic for Kosher certification. Personally I'd be inclined to look at bentonite first, as I've seen this used very effectively and it doesn't present so many challenges in terms of dietary suitability.
Apologies for the very long post, but I hope some of it is useful for you!