Crystal malt experiment - Attenuation test

After been told , like many of us (if you knew otherwise, I'm surprised), that sugar from crystal malts are highly un-fermentable, conducted this experiment to put this home-brewing myth under check.
The test consisted on brewing multiple 1gal batches, using crystal 10, 40 and 120L. Some batches were done using only crystal malt, some using only base malt (2row) and at last, batches with a mix of 50% crystal and 50% base malt. All batches were mashed at 155F and fermented at 70F (controlled chamber) with dry yeast S04, which was selected to be fast and promote good attenuation.

Here's the small mash-tun built just for the test:

All grains were milled 3 times to extract most sugars. For that reason, 1/4# of rice hulls were added to every grain bill to avoid stuck mashes.
Here's how the milled grain looked like.

Once mashed and properly boiled and cooled, wort was transferred to these small fermenters, oxygenated and pitched with re-hydrated yeast and set at my controlled fermentation chamber to exact 70 until fermentation was completed.

Daily gravity samples were taken and recorded, indicating when fermentation was completed, also making easier to visualize the fermentation profiles.

Overall, 18 individual 1 gal batches were made, as follow:
3 batches with 100% crystal 10L
3 batches with 100% crystal 40L
3 batches with 100% crystal 120L
3 batches with 100% 2 row base malt
2 batches with 50% crystal 10L and 50% 2 row base malt
2 batches with 50% crystal 40L and 50% 2 row base malt
2 batches with 50% crystal 120L and 50% 2 row base malt

The results were surprising. The table below shows some key data, recorded for all batches.

Click the table or charts to enlarge

Here's an iodine test showing starch coming from a crystal malt:

This experiment may not indicate exact values due to the reduced data points and accuracy of the tools used, but I hope it gives some light to what crystal malts do to our recipes.

Some key conclusions:

A)Crystal malt have sugars but starches are present that can be converted
B)The amount of sugars that one can extract from crystal malts would increase if mashed with a base malt since the starches will be converted. PPG showed to increase by about 20%, regardless of the kilning level of the crystal malt.
B)The sugars from crystal malts are VERY fermentable, contrary to what we thought. Fermentability will depend on multiple factors like:
-Steeping crystal malt alone will yield sugars that can be attenuated by 50% for crystal 10 and 40% for darker malts.
-Mashing crystal malts with base malts will yield sugars that are almost as fermentable as base malt itself. For the 50-50% rate used, sugars from crystal-10 malts were attenuated by 70% while crystal 40 and 120 by 52% minimums. For a 10% crystal to grist rate, I guess it could be treated just as a base malt, which means very fermentable.

From what I learned, some basic guidelines when using crystal malt would be:
1)If steeping crystal malts alone, expect lower PPG than when mashing it with a base malt. About 50% of the poits you get from the malt will be left to FG for light malts and 60% for darker malts
2)If mashing with a base malt, treat crystal just like a base malt, specially if using small amounts like 10 to 20% of the grain bill. So don't blame the crystal malt for a higher FG since most of its sugars will be fermented.
3)Regardless, crystal malt is not the best tool to add residual sugars to the final beer. Perhaps mashing at higher temp is the way to go, along with Lactose or Dextrin (that we believe is not fermentable. I may have to test that also).

For instance, if you use 2lb of crystal 40L to a 10lb grist recipe (20% rate), which is a lot, it would give you about 9 points for your OG and at least half of that will be fermented, leaving you with only 4.5 points of residual sugar from the crystal malt. For most recipes where only 10% of crystal is used, those residual points would drop to 2 or less, what doesn't give you any sweetness at all.