Vegetable Oil SVO WVO waste cooking oil as biofuel

[SunWizard]

I have a science project. Many are yearning for better tests than the HPT. To do this, we need to understand what is going on with the properties of our widely varying WVO. My theory is we need to test for FFA. The FFA test is ASTM D664 standard in the biodiesel world, and its set very low at <0.25% FFA, probably with a good reason since those ASTM chemists are smart. The german standard for VO fuel is <1% FFA which means new or lightly used WVO that has been handled well.

The biggest problem with FFA is it holds onto lots of water at a molecular level and it won't settle out. That water can be acidic if high FFA, causing corrosion of most metals. And it can turn into free water with a temp. cool down causing condensation in your storage or vehicle tank. FFA makes it harder to dewater. If you run some WVO with high water and FFA through a pump, it can make it much harder to dewater.

It also indicates the amount of polymerization already in your WVO, which can lead to deposits in your tank, IP, injectors. Its also a chain-reaction, so once it gets going it can increase very fast if there is oxygen available. Even after you dewatered the WVO, if high FFA it can make its own acidic suspended water while in storage from the oxidation reaction.

The trouble is there are few researchers working on WVO, they are mostly on Biodiesel. So we need to work together. What the Biodiesel folks have learned is very valuable to us since its similar in many ways. Most of the Biodiesel folks have already done these tests repeatedly and I will go ask them for help on this project.

First a definition so we don't go off (yet again) discussing what the meaning is:
I am using the term dissolved water as the water-in-oil testing industry uses: water so small its characterized by microscopic water dispersed and held tightly throughout the oil. Considered to be not harmful to our engines.

From http://www.nrel.gov/vehiclesan.../npbf/pdfs/46592.pdf
"Any water in fuels below the saturation level is soluble. At room temperature, the water saturation level of B100 is between 1,200 and 1,500 ppm" The saturation level is where we reach 100% limit of dissolved water, and it becomes suspended water and free water (many researchers call both free water since the only difference is the droplet size.) Just as with humidity, the saturation level varies with temp, see: Dissolved water and % saturation Maybe this means those in cold climates, and who have WVO that is close to the saturated dissolved level or high FFA (which means much more total water), should watch out for suspended water by doing another HPT after a large cool down.

The different tests for PPM will return widely different numbers, so please specify a test method when PPM is used. Such as PPM by KF, SB, or ASTM. In this post when I don't specify, its PPM by KF.

FFA:
WVO often has FFA (and mono and diglycerides MG & DG) up to 20% due to age. From:
http://www.biodieselmagazine.c....jsp?article_id=1708
"Even virgin soybean oil, typically a 1 percent FFA feedstock, can shoot up to 15 percent if left in a railcar during a week of hot summer days."
FFA are hygroscopic, meaning they attract and hold onto water at the molecular level as dissolved water.
"WVO has a surprisingly high percentage FFA. Yellow grease from local recyclers (central Illinois) runs in the 8-15% range in the winter and 12-20% in the summer." From www.istc.illinois.edu/tech/small-scale-biodiesel.pdf
Here is how it works, we start with a triglyceride which looks like a 3 pronged pitchfork. Then as it ages it gets damaged: when 1 FFA breaks off you have a diglyceride left, when 2 break off you get a monoglyceride. This is why measuring FFA is a good way to know how much mono and di-glycerides are there. This is why we can refer to the problem as FFA, while its really the MG and DG with the many nasty properties, but what they do to water is the big one.

"Mono- and diglycerides are amphoteric, meaning the alcohol “head” of the long chain likes to stay in water, while the “tail” of 18 or so carbon atoms prefer to remain in the fuel" From http://www.biodieselmagazine.c...le_id=1179&q=&page=2 which is interesting since they also say "There are other molecules, other dark Darth-Vader-type creatures that inhabit the fuel,” sterols, and "“What it does is amplifies what used to be just a residual trace amount of a [mono- and diglyceride], and they will suddenly form precipitates under certain conditions,” and clog filters. Biodiesel researchers have been researching to solve this issue for years, and I suspect we have it with WVO too since we have far higher FFA than Biodiesel is allowed to have.

A picture says a 1000 words, so I made this. Don't focus on the numbers as much as the theory, since the numbers will vary with type of WVO, and I may change it as we get more data points from more testers:


This could explain the large variation in ppm reported in samples that pass the HPT. HPT (with many limitations) lets you know in a crude way when you have crossed from dissolved water to harmful suspended water, while ppm (KF and SB) does not. Note this doesn't say anything about what levels of PPM or FFA I think are safe, everyone has to make that determination themselves. If we had a standards body make a limit, it would probably be like the Germans, a very small box. If you want to use WVO outside the box, it helps to know whats happening with your WVO so you can draw your own box (or triangle.)

It also points out a problem with ppm by KF or SB, unless you also measure FFA. For example, say I set a 500 ppm water standard. Some newer WVO may only hold 300 ppm saturated dissolved, so at 500 ppm I would have suspended water which the KF isn't warning me about. Old WVO may hold 800 ppm saturated dissolved (and pass a HPT,) so at 500 ppm I would be fine since it would remain dissolved with a 300 ppm margin. This renders KF a poor test by itself, and the old HPT would be needed for safety. Unless we also knew the FFA. If we had a chart for ppm x FFA, like I made above, we would have a much better idea of where we stand. This is similar to a problem even with the German standard of 750 PPM on new oil. I know from testing that 750 ppm in my new oil will fail a HPT badly.

FFA tests are fairly easy to do, see the Biodiesel forums for instructions, or send a sample to a standard testing lab. Many places sell simple FFA test strips made by 3M [/url] "To determine the percentage of FFA’s in your feedstock, simply dip the test strip into your oil and submerge for 2 seconds. Remove the strip and count the number of bands without any blue color." Sold here for $32 for 40 tests (better website). or
$24 per 40 here
The FFA tests for Biodiesel are done both before and after processing into Biodiesel. There are many ways published to reduce FFA, but they are much more work and science than most WVO users are used to, since you do things like titrate, add lye and water, stir, wait, makes soap, mistwash out the soap, test FFA again, etc. Its called caustic stripping. The trouble with it is even the most sophisticated mechanical separation (industrial Centrifuge) still results in at least 10% of the soaps remaining in the neutral oil. And we lose some portion of our WVO to the waste water (with 10% FFA we lose ~10% of our WVO.) Coorga quik n free claims to lower FFA much easier (20 minutes) than other methods.

A mistwash doesn't lower FFA much unless we also have water suspended acids in our WVO. A mini-mistwash is part of the TDS meter test I run for contaminants, which would tell us which batches its worthwhile to attempt to lower FFA by mistwash.

Here is research showing the huge variable range I have seen for PPM on WVO that passes a HPT, measured by SB and KF:
Report on Oxidation and Polymerization of Vegetable Oil
Note all samples were filtered, dewatered, HPT, and were from multiple forum members.
The sample that cgoodwin at frybrid submitted showed 777 PPM by KF. Others were 592,554,559, and 1140 PPM. In my own KF tests on the same batch old WVO before and after processing, 708 ppm failed HPT badly, and 545 passed with no bubbles.

http://www.frybrid.com/forum/s...=1825&highlight=ASTM
Post#26 he is processing 15,000 gallons of WVO a week and lots of lab tests:
"The samples I sent out to be spun down at 20,000 g's came back water free." , "The same sample tested via Carl Fischer came back with water at 2500 ppm. I think it is a FFA problem, and titration may not be a bad idea for wvo." This is an example of WVO that lab tested with 2500 PPM by KF of dissolved water! "The highest FFA counts (25%) are coming from places mixing animal fats with their WVO in the same barrel." and "I am certain that titration is a test that needs to be preformed an wvo."

Water tests:
The ASTM water test is D2709, and its done with a little centrifuge that spins these 100ml graduated vials at 1875 rpm after you add toluene as a reagent. This is a far different test than KF and SB, and we should look at its use, since it measures suspended and free water only (not dissolved.) It should correlate well to the HPT, unlike KF. This means we could set a very low standard for PPM ASTM, close to zero ppm, and varying FFA wouldn't change that standard. The KF and SB test includes dissolved water, which most people agree isn't harmful. So what good is using a test that includes that dissolved water if it varies hugely and isn't harmful? Its useful to tell us the % saturated, if we also tested FFA and made our own ppm x ffa chart. There are also testers and inline meters that will read % saturation directly, large producers may want to look at them. The trick is calibrating them. That would be done by making a chart just like I did above, and testing your FFA. And you also would need to develop a ppm x temp chart (I linked some above) since all the good meters are temp. compensated.

So as a summary for those asking for what tests to run to really know your WVO, ASTM D2709 water, D6304 KF water or SB test, ASTM D664 or test strips or titration for FFA, a HPT and a TDS meter test. With these we could develop a mini standard for WVO similar to ASTM for Biodiesel. I know a VO researcher who works at CSU (she asks me lots of questions), I may send her this and see if she can get funding for this kind of testing. Or maybe some big producers are still around like the person linked above who processes 15000 gal. per week, who was begging for some info like I provide in this thread.

We should test our FFA on a sample from the lowest point in a barrel we plan on using, or a well stirred barrel. As we should sample for our water tests like HPT or SB. Because water laden FFA are slightly heavier than the WVO and settle over a long enough period.

I know some biodiesel folks have run all 3 of these ASTM on their WVO feedstock, it would be great to see more data points from different batches of WVO to see whether it fits the above theory. Although a trick is that unless you know you barely passed the HPT or D2709 tests, you don't know how far below the saturated dissolved line you are.

[SunWizard]

Here is another chart almost like mine from a very respected VO researcher who published it in this report back in 2006 (see p.15-17):
Report on Oxidation and Polymerization of Vegetable Oil
.

Notice how the ppm numbers go even higher than mine and his description of the problem with FFA. I think we should test further his statement:
"The general consensus seems to be that, since FFA’s appear to be capable of complete combustion, they do not represent a particular threat. It is suggested however that this may be a case of interrogating the wrong suspect."

[SunWizard]

Some good news to report: I just won an auction at $45 for the lab CF needed to perform the ASTM D2709 water test that is used with Diesel and Biodiesel and discussed in the above post. Vintage hand crank, no longer made, with none of the electronics which make all the new ones real expensive. But it may wear out my arm or shoulder, its a big diameter crank radius to get that kind of speed. It needs 2 new glass vials, which I also just found at a bargain for $50 for 6. More are available on ebay if others want to play with them, but you need the rest of the centrifuge:
http://cgi.ebay.com/ws/eBayISAPI.dll?Vi ... K:MEWAX:IT

I have also gotten lots of offers of help, ideas, and data by people already.

[SunWizard]

FFA is also an indicator that severe breakdown is occurring in your VO which affects many properties, including polymers, aldehydes and ketone levels. Research shows that as FFA increases, viscosity increases due to the % of polymers increasing. Polymers create deposits and could explain why several people have reported injector "sticking" failure on high FFA VO.

USDA guidelines say restaurants should not exceed 2% FFA when cooking beef and poultry. I wonder how many restaurants even check?

[HoldOnTight]

SunWizard, My test strips should be shipping out tomorrow. It would be wise to talk about procedures for testing FFAs at this point.

Instructions say to dip the strip in hot oil. Is it necessary or would the result be evident if oil is room temp, but take longer. Testing temperature should probably be standardized... I can put a sample in a jar and test when it is at 75F with ease.

I use 2 methods of processing, Bag filter housing with Hot water heater and the other is an Simple CF with inline heater fed by a 60 gal cone bottom tank.

My filter setup and hoses probably holds about 5 gal of oil residual from the last processing batch. I can test the processed oil already in the fueling hose to understand the baseline condition of the processor. It was a blend from the 2 sources. I'm setting up a 2nd prefilter drum so I can keep the sources separate.

Thoughts?

My CF setup is currently inoperative while I reconfigure by setup in the garage. I'm also adding a dedicated high amp circuit for the CF & heaters...


BTW: I'm posting here because of the distraction caused by one individual's diatribe I prefer to avoid... I'm putting him on ignore, tomorrow.

[SunWizard]

Instructions say to dip the strip in hot oil. Is it necessary or would the result be evident if oil is room temp, but take longer. Testing temperature should probably be standardized... I can put a sample in a jar and test when it is at 75F with ease.

I don't know, to find out we will need to test it. Stir then split into 2 samples (tested cold and hot) from the same batch. Then we can compare notes on different batches. I have many old full cubies from different places sitting around I haven't processed so I can get many data points.

My filter setup and hoses probably holds about 5 gal of oil residual from the last processing batch. I can test the processed oil already in the fueling hose to understand the baseline condition of the processor.

That might not be a representative sample of the batch, do you still have some of it in storage?

I'm setting up a 2nd prefilter drum so I can keep the sources separate.

That would be useful if they are different FFA on your initial tests before processing.

[HoldOnTight]

We should test our FFA on a sample from the lowest point in a barrel we plan on using, or a well stirred barrel. As we should sample for our water tests like HPT or SB. Because water laden FFA are slightly heavier than the VO and settle over a long enough period.

Okay Sunwizard, I have test strips in hand and a 1/2 cup of preprocessed oil which was mixed before sampling. I am currently processing the batch, which is about 25 gal. because it is what I have in the Hot Water Heater already, and not enough to prefilter without holding up the processing line... I say that because I have about 50 gal. prefiltered and waiting for processing. On Monday, I'm leaving for a 1000 mi. (round trip) business trip, and I'll need about 50 gal of processed veg oil.

We need a consistent protocol to follow when taking measurements. Lets use my first small batch process to highlight and address the real-world issues that could affect the experiment.

We don't want to sample off the bottom or the top, IMO, due to variability of FFAs and settling time. So, IMO, we should sample once all the oil is mixed. This would take FFA settling time out of the equation. We want to minimize the variables for repeatability/comparison - a good thing.

This is my processing approach, as a means of addressing variables to be minimized:
I ran my pump to circulate the oil through the filter for a total of about 50 gal, or 2 filter passes, cold. This mixed the oil in the filter from the previous process session (Brazilian restaurant, 5 gal. put from fryer to cooling pot to cubbies) with the larger batch of oil collected from the Asian restaurant (25 gal dumped & held in my steel drum). I took a sample of this mix, which I am calling preprocessed, even though it has already been somewhat filtered...

I filter first because the contaminants are often polar molecules and the water is also polar. Filtering early on, IMO helps remove some dissolved water since it is bound to contaminants that can't pass the filter surface.

I then heated the oil to temp (140F), and ran the pump to filter for another 5 passes, maybe more. At this point the oil should be relatively free of >1 micron solid contaminants which may have been bound to water at the molecular level/dissolved.

I will allow settling now, which should be more effective without larger contaminants to hold water. At this point, the oil was filtered effectively to 1 micron. The filter is a size 2, which with hoses, so this is how I am estimating is about 5 gal. from the previous processing batch.

I will let the oil settle, then drain 10% (2.5 gal) to remove a majority of free and hopefully some dissolved water, then crackle test this oil. Why? Because you can never monitor too much to get insight to processing methods/steps. I'm going to compare it to the preprocessed sample which will also be crackle tested.

I'll pull a vacuum for about 2 hours once the oil is back up to 140F. By then, the oil should be bone dry. Then, if there is general consensus, I'll measure FFAs and crackle test the fully processed oil from a sampling port at the base of the water heater.

Think of this as a trial where we iron out the measurement protocol. We don't have to use the data..., but I could get some useful data from it!

I'd like to have the protocol ironed out before Friday, because I intend to process the next 50 gal in the pipeline, which came from the Asian restaurant's collection drum.

_ _ _

About this Asian client... They had no previous dumpster and I have no idea what they did with their oil prior to my arrangement. They were immediately on board to give me oil and they wanted to empty their fryer "this week!" I needed a drum quick, and not to disappoint the new supplier... I had no drum painted on the inside that wasn't already used for prefiltering and loaded with oil. So the only immediately available drum I converted for oil collection has a steel surface on the inside. I know, but it was put together as a short term solution to an immediate need. So here I am, measuring FFAs!

The drum sat for 2 of the hottest summer months between collections on the south side of the restaurant (nearly full sun all day, on a black drum) as the oil accumulated until collected. So the FFA test on this should be higher, maybe off the scale? This drum had about 1 gal of water that was also accumulated over 4 mo and 100 gal of oil. Most was drained after 75F prefiltering at 100 microns plus a 2nd stage (sheet) filter.

I have since heated the oil up to 140F, turned off the heat and it is settling now. The oil in the filter won't settle, even though insulated everywhere but the bottom, since it is in a closed loop with the Hot Water Heater (HWH) and isolated with valves on each end of the bag filter housing. Tomorrow evening, I'll drain the bottom 10% slowly, then I'll vacuum dry the water out. I'm guessing this small batch will be dried in an hour. I have a condenser jar frozen in ice, which I place on top of the HWH for vacuum drying. The tube runs out of the top of the HWH about 1 foot to the condenser and ~10 ft to the A/C vacuum pump.

When dry, the first indication is no moisture buildup on the short tube - looks exactly like the long tube. 2nd indication: I like to get to 26 Hg of vacuum with the oil around 130F or more and I know the oil is dry at this temp. Typically, I set a valve to let undried air enter the HWH at the base and I listen for a very slow bubbling. I'll aim for about 24" Hg initially. Once all the water is gone, the vacuum level should rise to at least 26" Hg.

Thoughts, comments, protocol suggestions?
Thanks,
HoldOnTight

PS. Sunwizard, Your comment in the first post about relative humidity in oil is not completely correct. First, everyone should understand that relative humidity exists in air as it does in oil, and the relationships for the variables are the same. This relative humidity of water in oil relationship has been known and documented since 1920. Relative humidity is a measure of the Dissolved water content relative to the Saturation of the oil/air. Relative humidity is dependent on not only temperature, as Sun pointed out, but also pressure (which is negative vacuum!). For this reason, I use vacuum dewatering to get free, emulsified and dissolved water out of the oil. Every crackle test comes out "bone dry" after processing with vacuum. FFA's as you suspect, may be the reason some are not able to pass the crackle test even though they CF's their oil. This has been my experience and it has driven me to add a vacuum processing protocol to include vacuum dewatering. But, that isn't to say I couldn't use a CF to process oil. If I know I have high FFAs then I can expect to get some bubbles with the crackle test, and I'll be comfortable knowing that it is okay too! Or, if I wanted to be really anal, I could collect the CF'd oil in a HWH, then vacuum dry it.

If the oil contains NO water (free, emulified or dissolved) then you could say that vacuum drying is another method to MEASURE all the water in oil, along with KF, SB... Incidentally, my last batch of oil from fryer to cubbies, measured 750 ppm after Heat/Settle/Drain alone, which didn't pass the crackle test prior to vacuum drying. After vacuum drying, the crackle test was absolutely boring...not the tinyiest bubble!

[SunWizard]

Sounds like good methods. I would also test the FFA before any processing, to see if the processing changes it. And one of the first things I want to check is how much variation there is between hot (300F) and cold tests on the same batch of VO.

I'm unsure which part of my statement about humidity you don't like, please give more info so I can make it better.

Was that 750ppm by KF? The KF test, which is considered the most accurate by far, measures the water by heating, applying a vacuum, collecting and condensing the evaporated water in toluene and measuring it very accurately. So you are correct about vacuum being a way to measure, if you could just collect the water and measure it accurately.

Asian places usually have very good VO, so I'd be surprised if its high FFA. Since most of my sources are good VO, I will probably want to get some of the low range <2% FFA test strips on my next order. For this testing I got the high range so I can test some old barrels I have had sitting, back from the days when I didn't have as good VO sources, which could give higher FFA and more obvious variation and results in testing.

[HoldOnTight]

Sunwizard, I took my measurement of the oil prior to processing, with the exception of prefiltering at 74F. This did remove some PHO. I tried to heat the oil to 375F, but the oil started to boil in the nuker. I checked the temp with my IR thermometer and it showed 211F. Realizing the water would boil, I decided to take the measurement at that temperature. The instructions indicated the test strip should be dipped in hot oil (directly in the fryer) for 2 seconds and the reading would be available immediately.

If the oil isn't at fryer temperature, then the reading takes longer to stabilize. At first, it looked like a reading >2%, but after 15-30 min of waiting, it indicated 2%<FFA<3> 8/3200 = .0025 or 2500 ppm.

Okay, here is what happened. Vacuum level isn't rising to where it was. I turned off the Vpump and I heard a very high pitch whistle. it was coming from the condensor lid! After emptying the water from the jar, I caused a leak at the fitting where it contacts the lid or I made a small leak a bigger leak! The vacuum level was 20 in when the water started boiling out. Just prior to draining the condensor a second time, it was at 24 in. and holding steady. After draining water the second time, the vacuum wouldn't rise above 22! Grrrr! I must have compromised the seal on the barb-lid junction when removing the vacuum hose from the barb. I knew the condensor jar was small and I had planned to replace it with a much taller jar to maximize the condensation opportunity and add more water holding capacity. I am going to add a valve as well so I don't lose the vacuum in the tank. It takes a while to get the vacuum up when the tank is half full of air! Nonetheless, this oil had so much more water than most of my oil!

This oil came from cubbies primarily, but that doesn't mean an occasional cubby doesn't have a way for water to enter. The cubbies are stored outside until the restaurant has enough to make a collection worthwhile... Some of the cubbies have the handle melted through slightly, so rain can enter. So, I've stopped dewatering. I would crackle test the oil, but it is late and I'll do that prior to decanting. I don't think I'll be dewatering any more. I believe this oil will still pass as nearly bone dry and it and the next batch of oil will be burned up in about a week. I'll also check the sample for FFAs at that time.

I've got the next condenser jar nearly ready. I've learned to shorten the down tube on this one, and I'm going to have a valve and 1 quick disconnect. So I'll do that tomorrow evening and process the next batch, about 50 gal. on Sunday.

Sun, I don't think the water content of this batch is reliable, due to the leak in the condensor. It would be helpful to know how long it takes for the cold reading to stableize to the point where it matches the hot reading, since we can't always get unprocessed oil up to fryer temperature for a test. So, can you take an FFA measurement from a processed sample of hot oil and then let it cool and take another FFA reading.

Cheers!

[SunWizard]

At first, it looked like a reading >2%, but after 15-30 min of waiting, it indicated 2%<FFA<3> 8/3200 = .0025 or 2500 ppm.

I don't understand the FFA or ppm part. Did you do a water test of some kind?

I am waiting on starting FFA tests since I also plan to run ASTM D2709, Sandy Brae, and HPT on each sample. I am still waiting for some of the equipment to arrive.

[zoochy]

water so small its characterized by microscopic water dispersed and held tightly throughout the oil. Considered to be not harmful to our engines.

Not to be dick but... if suspended water does not harm the engine then why dwell on the subject?

[SunWizard]

Not to be dick but... if suspended water does not harm the engine then why dwell on the subject?

Its dissolved water that's not considered harmful. The subject is knowing whats going on in your VO for lots of reasons spelled out in post#1. If you are near saturated dissolved, it can change into harmful suspended water with a temp. change. Or by humid air getting in your tank vent, then creating water drops by condensation. Or with high FFA you have polymerization and possibly large amounts of water.

There are 2 parts to this: 1 is testing our FFA. The 2nd is the hot pan test is very crude, and I want a better test. Before I decide which test, and what ppm I will use for my limit, I want to compare the different water tests, which I haven't seen anyone doing. To compare the different tests properly, and to understand what I am doing, you have to understand the info in post#1.

[John Galt]

I want to compare the different water tests, which I haven't seen anyone doing

KF titration, SB/Carbide manometer, vapor test, HPT...

my preference is the vapor test for ease of testing and cost

[HoldOnTight]

Instructions say to dip the strip in hot oil. Is it necessary or would the result be evident if oil is room temp, but take longer. Testing temperature should probably be standardized... I can put a sample in a jar and test when it is at 75F with ease.

I don't know, to find out we will need to test it. Stir then split into 2 samples (tested cold and hot) from the same batch. Then we can compare notes on different batches. I have many old full cubies from different places sitting around I haven't processed so I can get many data points.

My filter setup and hoses probably holds about 5 gal of oil residual from the last processing batch. I can test the processed oil already in the fueling hose to understand the baseline condition of the processor.

That might not be a representative sample of the batch, do you still have some of it in storage?

I'm setting up a 2nd prefilter drum so I can keep the sources separate.

That would be useful if they are different FFA on your initial tests before processing.

It would be nice if I could flush out 5 gal of oil from all the lines. But this is not a lab with a small filter. The residual may or may not skew the results. It really depends on the similarity of the initial conditions of each source. If they are widely different, they will impact each other. I don't have so much oil on hand to process twice from one source to prevent total influence of one source of oil on the other.

Also, I only use the crackle test equipment to test moisture level directly. Vacuum drying only shows when oil is really dry. So I have no quantitative measurements of water content. How to proceed with quantitative moisture levels?

[SunWizard]

Also, I only use the crackle test equipment to test moisture level directly. Vacuum drying only shows when oil is really dry. So I have no quantitative measurements of water content. How to proceed with quantitative moisture levels?

You need to do ASTM, SB, Carbide Manometer, or KF tests to get that kind of data. One test you could do with only a HPT is to add known amounts of water to a dried sample and see how much you add until you fail the HPT. For example, sample A, then test A+500 ppm, A+1000 ppm, etc.