33[hreg] Flame Tests on Papercrete/Fidobe
- Sep 13, 1999The following articles are from issue No.5 of Earth Quarterly. The
first article is on papercrete alone and the second includes tests on
COMBUSTION TESTINGON PAPERCRETE MIXES
by Robert Secrest
Over the past few months, I have been conducting some tests aimed
at getting data that can
be used to assess combustion hazards for various papercrete formulas.
This is a brief report on
the results to date.
I will save the details of my preparation methods for a full
report. It will suffice here
to say that care was taken to accurately weigh the components of the
mixtures and to prepare
and test the samples in a consistent manner. A propane torch flame held
at one point on the
sample for 60 seconds was the ignition standard used. This is much like
tests used by fire
PLAIN PAPER PULP
Samples made using nothing but newspaper pulp continued to flame
for 30 sec. to 1 min.
after removal of the ignition source. The flame spread was fairly slow.
After the flame
extinguished, the samples continued to smolder in a way similar to
charcoal in a barbecue.
Some samples self extinguished after a time. Others were completely
camumed, leaving a residue
of white ash and unburned carbon.
PAPER PULP with BORIC ACID
As a theater person, I have used boric acid as a fire retardant on
scenery. I felt it was
worth testing in this context. I added a quantity of boric acid to some
samples of plain paper
at 20% boric acid by weight to the dry paper. These samples did not
support flame at all after
the ignition source was removed. They continued to smolder and produce
smoke for about a minute
after, but soon self extinguished completely. Boric acid is clearly
effective as a fire
retardant, but represents an added step and cost factor (even when
purchased as common roach
powder). I have not, as yet, established the minimum effective ratio for
this additive. It is
possible that a saturated solution applied to the surface like paint
would also be effective.
MIXTURES CONTAINING PORTLAND CEMENT
In spite of being the most popular formulae for paper based
blocks, the cement mixtures
are the most troubling in terms of combustion. Samples containing
moderate amounts of cement
(50% and 100% by weight to dry paper) all exhibited the intense
flameless combustion, which I
refer to as "slow burn." Unlike ordinary combustion, which produces
smoke and a visible flame,
the "slow bum" produces a great deal of heat, (and I suspect Carbon
Monoxide) while showing no
flame and little smoke. The process is tenacious and will continue to
run through every sample
until it is completely reduced to a light brown ash. Quenching the
sample in water is about the
only way to extinguish this reaction. Samples containing a portion of
sand equal in weight to
the cement showed no difference with respect to the bum. I have not yet
tried samples with high
ratios of cement (2:1 and 4:1). These levels, while of interest to the
tests, would probably
represent impractically high ratios in block production.
I also made samples containing an equal quantity of Portland cement
and paper (by weight),
to which I added 20% boric acid. In this case, the samples did burn,
despite the addition of
One should bare in mind that, in spite of this disturbing evidence
on cement formulae, the
ignition method used represents a rather intense starting energy level.
The risk of ignition in
our applications may be acceptable.
This testing is to be ongoing and I hope to have a wider range of
results to report on
later this year. I am quite interested in clay/paper mixtures, which
show much promise.
I am happy to accept questions, comments, and suggestions by e-mail
Thanks to the Solbergs for including my report in EQ.
After we performed the strength tests at NMSU (see page 21) we had
plenty of broken
fragments left over, so I performed some flame tests on them. All the
listed on page 21 burned. These had a nonflammable (cement and sand)
content ranging from 44%
(the roof panel mix) to 65% (the formula for large tow mixer). Pure
paper pulp produced a lot
of smoke; of the papercrete samples, the higher the cement content, the
less smoke was produced.
I also flame tested my paper adobe test bricks. The 1:1
(dirt:paper; 50% nonflammable)
and 2:1 (67% nonflammable) bricks burned. The other bricks, starting
with 3:1 (75% nonflammable)
and ending with 5:1 (83% nonflammable) did not burn.
Flammability evidently depends on the ratio of flammable to
nonflammable material. For a
nonflammable formula, the critical level of nonflammable material is
between 67-75%. A little
bit of further testing should allow us to accurately pin down this
---Gordon Solberg, Editor