Bill, I have done some research into EPS processing as well, and in many cases I do agree with you. In fact, to build a state of the art EPS facility currently does cost about $2 mil USD, but what you get for that is a very mechanized plant that can produce millions of cubic feet of Expanded Polystyrene per month. And when you look at it as a capital investment, it really isn't that extreme - amortized over 30 years it works out to less than $200 per day. For a 3 shift, 24/7 plant, that is rather inconsequential.
It also buys you systems that are much more efficient in terms of energy usage and quality control. This is mainly due to vacuum assisted molding techniques and an array of hi-tech sensors that provide feedback to the computer controlled automation. All of this provides some degree of advantage over other fabricators that are still using much more manual processes. And, historically, it has been difficult to locate qualified personnel to operate older less automated machinery, that is much less true with a new $2 mil facility.
That said, you do bring up some other issues. Like locating unexpanded EPS beads. It is possible to find them, but the price is very directly tied to the cost of oil. But it is highly unlikely that existing fabricators have any real advantage (with the possible exception of volume discounts and established releationships with suppliers). But any agreement to purchase a commodity that is so directly dependent on oil prices necessarily has cost adjustments that are tied to the feedstock price. I have studied EPS fabrication a bit myself, and have been able to locate unexpanded beads for single pound retail quantities at about $3.5/pound (it used to be $2.50/# 6 to 8 months ago).
Comparing that with EPS board that you can buy at Lowes or HomeDepot that is 2" x 4' x 8' and 1# density for $21.5+ per board, you arrive at a retail price on the order of $4.00/#. Both of those are both retail prices, and very small quantity prices as well, so they are pretty comparable. That price is also pretty close to prices available from on-line suppliers as well. So, locating a wholesale supplier (and getting prices closer to 40% to 50% of retail) shouldn't be difficult. And to them, every customer is just another demand for their product.
All this being said, the only real advantage that any EPS molder can offer is being local to the demand. Unexpanded EPS beads can be shipped very easily. When you expand them to 40 times the unexpanded size, the shipping costs can become enormous - not because of weight, but volume. So, the real question to be asking one's self is: what is the local demand for formed EPS? And, how much local supply is available? The answers to those questions could propel one into a profitable venture . . . or not.
But, now, getting to my real point. . . I assume that most here are interested in EPS for SIPs fabrication. While EPS may be less expensive than PU (polyurethane foam), why would anyone want to use EPS when PU is a better alternative. Consider this, 2 part unexpanded PU resin can be purchased for a retail price of about $5/expanded cubic foot (2# density) vs. $3.5 to $4.0/cubic foot for EPS (1# density) either expanded or unexpanded. So, PU sounds on the surface to be more expensive, although more dense.
But, with EPS, after it is expanded, it then has to be glued to the skins (OSB or fiber cement board, or plywood, or whatever) of the SIP. The glues necessary to do are not an inconsequential expense either. Take a cheap (and I mean cheap) polyester based glue with a coverage of around 2mil to 3mil. Current retail prices in drum quantities of that glue amount to $10 per 4x8 eps panel. That plus the fact that the expanded EPS foam has to be warehoused, and then handled by machinery to cut it, and then apply the glue, and position it in the panel and then press the panel until the glue cures all adds up to the price of EPS being much closer to PU. That is due to the fact that PU is mixed and injected into a SIP in one operation, and expands within the panel, and is its own 'glue'.
So, the cost difference dissapears when you factor in the handling and glueing expenses. Then, I read on Thermasave's website that PU foam emits cyanide gas when it burns. Any of you who haven't encountered Hoot Haddock, should check out his website (I think it is www.thermasave.com). His name is quite appropriate, because he is quite a "hoot". But, that statement of his is highly misleading, and alarmist (which does cast a bit of a shadow on Hoot's reputation). PU foam does burn (just like EPS does) and can produce trace amounts of hydrogen-cyanide gas when it burns. But the amount produced might be enough to give you a headache, but not enough to do any real damage - it if did, then the fire would certainly get you first.
When you factor in that PU and EPS basically cost the same to place into SIP sandwiches, then the next consideration is that PU is much more efficient thermal insulation than EPS . . . by a factor of about 2:1. That should make it easier to sell PU than EPS. The one last advantage to PU over EPS that I will mention here is that PU panel manufacturers like Murus (www.murus.com) have discovered that by using PU foam injected between the skins of the panel don't require splines to interconnect adjacent panels.
I presume that they do nail down to a continuous sill plate, and at the top to continuous top plates, but adjacent panels can be connected by using camlocks that are simply embedded into the foam. The adhesion of the PU foam to the skins and to the "wings" and bodies of the camlocks is strong enough that no other splines are required. They even claim that the camlock connection is stronger than nailed (and foamed / glued) in splines.
So. . . before anyone decides to invest money in a plant to manufacture more EPS blocks, they should at least consider PU in a foamed-in-place processes instead.
Dick