Intoxication Analysis

Science Summaries

1. Semiconductor Breathalyzers
AlcoMate (AL Series)
BACtrack (S30, S50, S70)
Q3 Innovations (AlcoHAWK)

2. Fuel-Cell Breathalyzers
AlcoMate (AL Series)
BACtrack (S70 and S80)
CMI (Intoxilyzer Series)
Draeger (AlcoTest Series)
Intoximeters (Alco-Sensor Series)
Lifeloc (FC Series)
Q3 Innovations (AlcoHAWK Series)

3. Measurement Considerations
4. Sources of Breathalyzer Error
5. Breathalyzer Conclusions

Below, I present an alcohol breathalyzer comparison summary and buying guide. If you are concerned about alcohol intoxication, you may want to purchase a handheld digital breathalyzer. These devices ESTIMATE your BAC (Blood Alcohol Concentration), by measuring the amount of alcohol emitted from your breath (BrAC).

Are you planning to buy a breathalyzer? Here is a list of breathalyzers currently being sold. I have listed the manufacturer, model #, power source, readout type and recent price, to aid in your purchase.

There are two main types of technology that are used to determine BA level. The older, less accurate technology is called "semiconductor module sensor". The newer, more accurate technology is called "fuel-cell sensor". Below, I offer a summary of the two types of devices.

top of page

Manufacturer /

Model #

Model Description /

Available From

AlcoMate
(AK Solutions)

Palisades Park, NJ

Alcoscan AL2500

Single button
No mouthpiece
Service required after every 200 tests

Alcomate - $ 60

AL2500 Elite

The Alcoscan AL2500 PLUS mouthpiece
Service required after every 200 tests

Alcomate - $ 80
Amazon - $ 55

Prestige AL6000

Must replace sensor after 300 tests

Alcomate - $ 130
Amazon - $ 124
Costco - $ 115

Premium AL7000

Sensor modules should be replaced every 6 - 12 months.

Alcomate - $ 190
Amazon - $ 158

BACtrack

San Francisco, CA

Described as "trusted" and "advanced", yet they are inferior to fuel-cell technology
Sensor accuracy is specified at .02% BAC. It SHOULD BE specified at .08% - don't you think? Who cares how accurate it is at .02%, where it doesn't matter. The reason they specify this way, is because the accuracy is FAR LESS at 0.08%, where it really matters!

S30
BTS30

BACtrack Select S30

Target has 3 models with SAME NAME - !

Amazon - $ 60
Target - $ 50-60
Off. Max - $ 50

S50

BACtrack S50

Amazon - $ 60
Costco - $ 50
Target - $ 60

S70

BACtrack Select S70

Amazon - $ 100
Best Buy - $ 80
Target - $ 100

Q3 Innovations

Independance, IA

Sensor accuracy is specified at .02% BAC. It SHOULD BE specified at .08% - don't you think? Who cares how accurate it is at .02%, where it doesn't matter. The reason they specify this way, is because the accuracy is FAR LESS at 0.08%, where it really matters!

AlcoHAWK Slim

Q31-2800

Amazon - $ 40
Best Buy - $ 60
CVS - $ 50
Walgreens - $ 50

AlcoHAWK Precision

AlcoHAWK Precision - Q31-3000
Single button

Amazon - $ 69
Best Buy - $ 80
Target - $ 70

AlcoHAWK Pro

AlcoHAWK Pro - Q31-11000

Best Buy - $ 150
Target - $ 140

top of page

Model #

AlcoMate
(AK Solutions)

Palisades Park, NJ

AccuCell AL9000

Cumulative test counter
Calibration required after 1,000 tests

Alcomate - $ 300

BACtrack

San Francisco, CA

S75 Pro

Appearance is another difference; the S75 Pro has been designed to reduce unnecessary size and weight found in the S80, and has a front-facing mouthpiece.

Amazon - $ 152
Best Buy - $ 150

S80 Select
S80 Pro

"The S80 Pro's larger fuel cell sensor will provide readings that are slightly more accurate than the S75."
Define "slightly more"
I have reviewed this unit below !

Amazon - $ 194
Costco - $ 150
Target - $ 200

CMI

Owensboro, KY

Every device has a different range - .400, .450, .500, .600. Can't they make up their mind?
Intoxilyzer Error

Intoxilyzer 240

POOR readout - cannot be read, even on the COMPANY'S OWN PDF brochure.
Accuracy is specified at ONLY 2 points!
Reset Time: NOT specified

Rep. only
(High-priced)

Intoxilyzer 300

The brochure states "4-digit display". This is CLEARLY incorrect. Apparently, one digit is invisible. One can only imagine other, more subtle spec errors.
Reset Time: 2 minutes to reset after positive test.

Rep. only
(High-priced)

Intoxilyzer 400 PA

Reset Time: 2 minutes to reset after positive test.
Range: NOT specified

Rep. only
(High-priced)

Intoxilyzer 5000 EN

Reading doesn't start at zero, but at 0.005% !?!
Reset Time: NOT specified

Rep. only
(High-priced)

Intoxilyzer 8000

Reset Time: NOT specified
" . . with a separate DSP processor for signal conditioning."
Signal conditioning? - what kind of hocus-pocus is this? Garbage in, massaged result out?

Rep. only
(High-priced)

top of page

Draeger

Durango, CO
AlcoTest Error
Problems with Drager Breathalyzers

Alcotest 6510

Retains data from last 10 tests
Type approval in 2005

Ace - $ 775

Alcotest 6810
(extended)

Retains data from last 250 tests
Type approval in 2008

Ace - $ 1,014

Alcotest 7410

This unit is used by the Santa Cruz County Highway Patrol at Dominican Hospital, but is no longer sold, and has not been sold for some time.

Type approval in 1993 -
drinkdriving.org/breathalysers
This device type is 17 years old. Available online USED for $398

Why is the County of Santa Cruz using an obsolete device, designed over 17 years ago?

The most accurate, modern breathalyzer devices have 3 significant digits to the right of the decimal point.
The Alcotest 7410 presents data with ONLY 2 significant digits to the right of the decimal point, as can be seen in the printout on the left.

Recent results on the left from Santa Cruz, CA show a printout of 0.09. Notice that it is NOT 0.090. The next lower reading is 0.08 - the third significant digit is NOT PRESENT.

This raises several issues,
1. Is the device incapable of producing 3-digit accuracy? Yes, by definition!
2. Just how does the rounding occur? Up or down?
3. If you are 0.085%, does the device give you a 0.08%?
4. If you are 0.075%, does the device give you a 0.08%?
5. Note that the difference between 0.08 and 0.09 is 12.5 % (1/8), which means that JUST THE READOUT ACCURACY at that BAC level is 12.5%! The measurement accuracy is not germane, since the readout display is far less precise, and establishes the limitations of the device.
6. I suggest that the lack of a third digit PURPOSEFULLY obscures the lack of the repeatability and accuracy of the device. What kind of "number massage" goes on inside the software? Has anyone ever audited this?

Alcotest 7510

Retains data from last 5,000 tests

Ace - $ 1,782

Alcotest 9510

Dual-sensor, touch screen
3 USB ports

-

Intoximeters, Inc.

St. Louis, MO

Alco-Sensor III

Latest test recall

Alcopro - $ 519
Alcopro - $ 350 used

Alco-Sensor IV

"Automatic calibration" - WHAT?
Used by Santa Crux County for field testing (PAS - Preliminary Alcohol Screening)

Issues with the Alco Sensor IV

Alcopro - $ 740
Alcopro - $ 395 used

Lifeloc Tech.

Wheat Ridge, CO

FC 10

5 buttons
Accuracy is + or - 5% above 0.10 BAC
Retains data from last test

BSS - $ 480

FC 20

Accuracy is + or - 5% above 0.10 BAC
Internal real-time clock
Retains data from last 500 tests

BSS - $ 785

Q3 Innovations

Independance, IA

Target chooses NOT to specify Model Numbers - It is very difficult to figure out which model is which.

AlcoHAWK PT500

3 or 4 digit - they won't say, or show
3 buttons
Sensor accuracy is stated at 0.04% BAC - who cares?
What is the accuracy at 0.08% BAC?
I suspect that this model has a lousy display, since they will NOT SHOW it, or even describe it!

Amazon - $ 152
Target - $ 200

NOTES:
1. ACCURACY - Number of digits - Note that there are 2-digit, 3-digit, and 4-digit models. However, in actuality, there are only two models, a 2-digit and a 3-digit, because the first digit, which is the digit BEFORE the decimal point, is MEANINGLESS, because it is ALWAYS ZERO. Some "3-digit" models are actually "2-digit" models (the 7410), and ALL "4-digit" models are in reality "3-digit" models. A device having 3 digits to the right of the decimal point CAN BE TEN TIMES more accurate than a device having only 2 digits to the right of the decimal point. Do 2-digit devices round up or down? Just how do they round off?

2. BSS = Breathalyzer Super Store. Ace = ace-breathalyzer.com

3. Other types of breathalyzers are available, but the devices listed above are among the most popular.

4. Many of these breathalyzers can be purchased from mainstream retail vendors, including Amazon, Best Buy, Costco, Office Max and Target - but ONLY online. No large retailer stocks them. The Amazon price will change every time you look.

top of page

160

320- 900

100

160

160

top of page  

.
BAC Level Variations

3.1

ALCOTEST PROBLEMS

From http://www.totaldui.com/news/articles/breathalyzer/source-code.aspx
and . . http://www.duiblog.com/2007/09/04/secret-breathalyzer-software-finally-revealed/

"Also, the Draeger AlcoTest device takes an initial airflow reading at startup to use as a baseline measurement. However, there is no testing or accounting for changes in air flow that might be different from this arbitrary baseline. Thus, a range of possible values may be inaccurate simply because the initial reading may discount it.

The values for the readings, which have the possibility of a range between 0 and 4095, are divided by 256 within the machine function, and so there are only 16 possible reading grades, not nearly the level of precision required by a breath test device."
---------------

1. The Drager AlcoTest is used by Santa Cruz County, as their "evidence standard". In fact, they employ a Model 7410, which is an older model similar to the Model 7110 device discussed here, and is no longer offered for sale.

2. The Draeger AlcoTest 7110 specifies a range of 0.000 to 0.500.
    The Draeger AlcoTest 7410 specifies a range of 0.000 t0 0.400.
I have drawn the corresponding scale above for the 7110.


From http://www.newjerseycriminaldefenseblog.com/2007/12/alcotest_not_much_improvement.html

"Because Alcotest also relies on breath testing, we can expect many of the same problems. Much worse, however, are the serious problems a court-appointed software testing firm found with the device's software. In short, Base One Technologies found the software was so poorly tested that it doesn't meet federal government standards; the FAA would not be legally permitted to use it for alcohol-testing of commercial pilots."

3.2

BREATHALYZERS - WIKI

From http://en.wikipedia.org/wiki/Breathalyzer

Common Sources of Error
"Research indicates that breath tests can vary at least 15% from actual blood alcohol concentration. An estimated 23% of individuals tested will have a BAC reading higher than their true BAC."

(Conclusion: Since the difference between 0.08 and 0.09 is 12%,
this is well within the stated 15% cumulative error stated above.)

Interfering Compounds
"Some breath analysis machines assume a hematocrit (cell volume of blood) of 47%. However, hematocrit values range from 42 to 52% in men and from 37 to 47% in women. A person with a lower hematocrit will have a falsely high BAC reading."

(This can clearly be a significant source of error.)

Homeostatic Variables
". . there is a 2100-to-1 ratio of alcohol in blood to alcohol in breath. However, this assumed partition ratio varies from 1300:1 to 3100:1 or wider among individuals and within a given individual over time. Assuming a true (and US legal) blood-alcohol concentration of .07%, for example, a person with a partition ratio of 1500:1 would have a breath test reading of .10%‚ over the legal limit."

3.3

BAC - BLOOD ALCOHOL CONTENT - WIKI

From http://en.wikipedia.org/wiki/Blood_alcohol_content

Test Assumptions - The Partition Ratio
"Blood alcohol tests assume the individual being tested is average in various ways. For example, on average the ratio of BAC to breath alcohol content (the partition ratio) is 2100 to 1. In other words, there are 2100 parts of alcohol in the blood for every part in the breath. However, the actual ratio in any given individual can vary from 1300:1 to 3100:1, or even more widely. This ratio varies not only from person to person, but within one person from moment to moment. Thus a person with a true blood alcohol level of .08 but a partition ratio of 1700:1 at the time of testing would have a .10 reading on a Breathalyzer calibrated for the average 2100:1 ratio."
Also, from http://www.avvo.com/legal-guides/ugc/using-the-variableness-of-the-2100-to-1-partition-ratio-to-attack-breath-tests
and from Denver - http://www.denverdui.com/drunk_driving_defense/blood_partition_ratio.html
and more - Google search for partition ratio blood test

Partition Ratio on Google

(Conclusion: This error ALONE can explain why
a 0.08 level can be read as a 0.10 BAC level.)

Metabolism and Excretion
"Rate of detoxification of alcohol can also be slowed by certain drugs which interfere with the action of alcohol dehydrogenases, notably aspirin"

(Conclusion: Daily aspirin therapy can alter the rate of
elimination, resulting in abnormally high BAC readings.)

Retrograde Extrapolation
"The rate of elimination in the average person is commonly estimated at .015 to .020 percent per hour, although again this can vary from person to person and in a given person from one moment to another.

3.4

SCIENTISTS: ASPIRIN RAISES BLOOD-ALCOHOL

From http://www.duiblog.com/2006/02/15/scientists-aspirin-raises-blood-alcohol/

"In vitro, aspirin clearly decreased the activity of gastric alcohol dehydrogenase in human subjects"

60

370-740

60

BROAD, QUESTIONABLE PHYSICAL ASSUMPTIONS
1. Hematocrit level is assumed (can be changed by disease and/ or medication)
2. Partition Ratio is assumed - it can vary considerably by individual, and time
3. Body Type
4. Gender variation

TIME ELEMENT
-- Delay up to legal measurement - can result in an increased level or a decreased level - the level should be dropping if there was no recent consumption

AIR SAMPLE INTEGRITY
1. Top air vs. deep air - hyperventilation vs. slow breathing
2. Mouth alcohol - mouth wash, cough syrup
3. GERD - stomach alcohol is not breath alcohol - fat people learn to supress obvious burps

EQUIPMENT CALIBRATION
-- When was the last calibration?
-- Credentials of the calibration company? The technician?

MEDICAL COMPOUNDS
-- Aspirin therapy effects - hematocrit reduction
-- Other medications

MEDICAL ANOMALY
-- Partition Ratio is dramatically different than the average
-- Hematocrit level - constantly varies dramatically in some individuals

Trombetta Advisement
Failure to Advise of Breath Sample Destruction and Right to Independent Test

A person who chooses to submit to a breath test must be advised, before or after the test, that the breath-testing equipment does not retain the breath sample(s) for later retesting, and that he may submit to a urine or blood test at no additional cost to him. CVC §23614. See also, CVC §23612(d) (right to independent test).

The failure of an officer to so admonish the accused does not trigger suppression of the breath test results from the evidence at trial. CVC ¬ß23614(d). However, the fact of the omission is admissible at trial and the defense is entitled to a jury instruction‚ advising the jury of the provisions of [CVC ¬ß23614, formerly ¬ß¬ß23157.5 and 13353.5], and permitting the jury to consider the officer's failure to advise the defendant of his rights as a factor bearing on the credibility of the officer and the accuracy of the breath test administered by the officer or at his direction. People v. Alvarado (1986) 181 Cal.App.3d Supp 1.

top of page

CONCLUSION -
Legal BAC Threshold Measurement Is Mighty Iffy

Anyone who produces a 0.24 BrAC is clearly whacked. I have no issue with this.

HOWEVER, when used to determine a BAC of 0.08, there are so many variables associated with these devices, that ANY DETERMINATION of 0.08% BAC is ARBITRARY, and INEQUITABLE, at best, for the following reasons. A composite error of 15% is easily achievable.

Subject Variation
- Medical conditions - hematocrit level, partition ratio, GERD, aspirin therapy
- Mouth alcohol, food ingestion, smoking within last 20 minutes
- Body type - gender variation - NOT part of the equation
- Breathing technique - minimizing deep lung air in the sample - surreptitiously deeply expel, then deeply inhale
- People with huge lung capacity at great advantage - can expel mostly upper lung air - another inequity

Measurement Variation
- Device variation - calibration, no consideration of recent contamination, can be knocked out of kilter
- Improper device spec'ing, NO SPECIFICATION over the RANGE of readings ENCOUNTERED
- Operator error - modified breathing technique not detected (I will run another test specifically for this)

Semantics
BAC is frequently used incorrectly, when BrAC is the proper term. This omits, and glosses over, the important fraught-with-error conversion process, with its various broad assumptions, especially concerning partition ratio.

Device Design
Bogus auto-repeat software to imply greater device accuracy, such as used by Weight-Watcher scales
Weight-Watcher scales use a rather underhanded method of readout, to imply that the scales are more accurate than they actually are. If the person being weighed gets back on the scales within 30 seconds, the machine does NOT produce a new reading - instead, it merely repeats the last reading, to make the user feel that the scales are REALLY accurate. However, If you wait for 1 minute, you will get a new reading (because it might be somebody else). Do manufacturers of breathalyzers think like this? They need every ounce of credibility they can get. Could they? Yes. Would they? Who has checked their software?

BREATHALYZER TESTS ARE AN INACCURATE MEASUREMENT OF BLOOD ALCOHOL CONTENT

top of page

50

500-700

50

DISCLAIMER
Although the author has tried to be as accurate as possible, errors are possible.
I could not find a concise Breathalyzer summary like the above. So, I made this alcohol tester comparison chart, to help myself understand the technology and pricing available. I hope that it helps you.

Junk Science plus Junk Cops = Contrived Conviction


RELATED PAGES -

Alcohol-Breathalyzers Compared

BACtrack S80 Breathalyzer Review

Santa Cruz, CA Criminal Defense Attorneys

.

top of page