California History of Defendant Challenges to Breath Test Results in Per Se DUI Cases 

      Before the California Supreme Court opinion decision in People v. Bransford

(1994) 8 Cal.4th 885, (“Bransford Court”) anyone arrested for driving under the influence (“DUI”) on what is known as a “Per Se” DUI charge pursuant to California Vehicle Code §23152(b) for driving with a 0.08 blood alcohol level and then given a breath test, was entitled to raise as a defense that the breath test result did not accurately reflect the driver’s true blood alcohol level at the time of driving.¹ This right was set forth in several cases, including People v. Cortes (1989) 214 Cal.App.3d Supp. 12, 17-18; People v. Lepine (1989) 215 Cal.App.3d 91; 95-96, People v. Thompson (1989) 215 Cal.App.3d Supp. 7, 11-12; and People v. McDonald (1988) 206 Cal.App.3d 877, 880.  Prior opinions holding to the contrary were People v. Herst (1987) 197 Cal.App.3d.Supp. 1, 3-4; People v. Gineris (1984) 162 Cal.App.3d Supp. 18, 24-25; and People v. Pritchard (1984) 162 Cal.App.3d Supp. 13, 16-17. Those courts have now repudiated those prior decisions and acknowledge that it was wrongfully decided.  

      Other states have also permitted challenges to the breath test blood alcohol level evidence, and those legislative bodies have responded by enacting statutes that that specifically outlaw driving with a specific and certain blood alcohol level by providing defendants with little alternative but to plead guilty   Those states having court approval of these statutes are:  Wisconsin, State v. McManus (1989) 447 N.W.2d 654, Washington, State v. Brayman (1988) 751 P.2d 294, and Illinois People v. Gustafson (1990) 551 N.E.2d 826.  Each court held that 0.10 grams of alcohol per 210 liters of breath was sufficient evidence presented to conclude that there was a rational relationship between alcohol on the breath and driving ability, and the state’s prohibition to this conduct was proper. In particular, the Wisconsin and Washington courts both disagreed with the defendants positions that that the statute created a mandatory presumption of blood alcohol level based upon a specified breath alcohol level because those statutes did not address blood and only made driving with a certain amount of alcohol on the breath illegal.

Most breath alcohol test machines are characterized as either a Preliminary Alcohol Sensor (PAS) or an Evidential Breath Testing device (EBT). The PAS devices are either pactive.  Passive devices measure the air space around the suspect, and do not require the subject to deliver a breath sample directly into the device. Active devices require the subject to blow directly into the device. and passive. Originally, PAS devices were only to assist the officer in determining if alcohol was present. The machines most commonly used for evidential purposes are stationary devices that are maintained in a controlled environment, Most of the modern EBT machines employ infrared technology or a combination of fuel cell and infrared. The California Department of Justice uses a fuel cell device, the Draeger 7410, as both a PAS and EBT. Regardless of the machine employed, the machine must be listed (without an asterisk) on the Federal Conforming Products List (CPL) if it is to be used for DOT (work place) testing under the Omnibus Act of 1994 [49 CFR part 40]. The Conforming Products List identifies devices that have been tested and shown to conform to the model specifications, as determined by the Federal Government. [NHTSA, 2002, 67 FR 62091.] 

      The stationary devices used by law enforcement include the Alco-Sensor® RBT and FST; the Intoximeter® EC/IR and EC/IR II; the Draeger Alcotest® 7110 MKIII-C and Draeger Alcotest® 9510; the Draeger Alcotest 7410 ^Plus; the Intoxilyzer 5000 and 8000; the BAC DataMaster, and the DataMaster CDM and DataMaster DMT;  Other miscellaneous devices include the Breathalyzer® Model 1000 and Models 900, 900A; the NPAS® passive alcohol sensor and the Life-Loc® PBA 2000; the Alcolmeter pocket model; and the Swedish Evidenzer.   

      The known issues with these machines include capturing a breath sample that can be reliably tested at a later time, confirm the original breath test result, and accuracy or calibration checks. A forensic scientists concern should always be to use a reliable testing device such that informed opinions can be stated in a court of law.  In reviewing individual cases, the scientist should evaluate and consider: The technology employed by the device; any unusual circumstances specific to the subject (e.g. exposure to interfering substances); the maintenance records of the machine, and whether the repair was effective; the calibration records of the machine, and the frequency of such calibration; checks on the internal calibration of the machine via wet bath simulator or dry gas; the performance of the machine near the time of the case test, as evidenced by the analysis of other subjects; the administration of the test, notably the 15 or 20 minute observation period prior to testing; the scientific community’s opinion (and the basis of the opinion) regarding: taking two tests; the Time gap between the two tests; the inherent margin of error in the machine, and the margin o error with human breath testing; and physiological aspects, which lead to differences in BrAC vs. BAC.


 It is the specific goal of scientits and the developers of breath test machines that the results are accurate, reliable and produce no false positives.  Reliability is a function of the administration of the test, the technology employed by the machine, physiological variables, and accuracy and precision of the machine with not only a simulator, but human breath. Precision is the ability to reach the same result, time and time again. Accuracy is how close the result is to the true value. Each machine is internally calibrated and each laboratory has different protocols for determination of accuracy, with the internal calibration is performed by the manufacturer or a factory trained technician. Experts disagree on the actual error in breath testing, so the issue is inconclusive.  

      Proper administrative factors may affect the breath test outcome, and must follow not only the manufacturer’s guidelines but also the guidelines in the prevalent scientific community.  Some States truncate the actual raw data on the breath card or digital display from three to two digits BrAC. This does a disservice to the accuracy of the results. The safeguards that may be put in place are: A 15-20 minute deprivation period prior to testing; duplicate analysis of specimens; a control test with each subject test; and having blank tests before each breath collection to ensure no carry-over of a previous test.  

      The 15-20 minute alcohol deprivation period is a universal requirement necessary to assure the validity and the legal admissibility and acceptability of the results of breath-alcohol analysis when carried out for forensic purposes. Duplicate tests are universally performed in blood and urine alcohol testing, and this standard should also apply to breath samples. They are use is over half of the states, with a premise that only with duplicate tests can you help rule out random error. Collecting one specimen and analyzing it twice is not considered duplicate testing, since the second analysis is not from the collection of an independent second breath. However, two measurements of the same sample is called replicate testing. Replicate testing is a useful tool for troubleshooting unstable electronics and/or detecting radio frequency interference. Moreover, the human variables on the accuracy of breath testing make duplicate testing mandatory.  A control group in the test allows for the reduction of error, self-deception and bias.  In an air blank test, air is pumped through the chamber to clear out the chambers contents prior to each breath test, between each breath/calibrator sampling, and after each breath test.The use of a air blank test is to clean out the chamber and to prove that a previous sample is not contaminating the current subject’s test. Carryover of a previous subject’s alcohol-laden breath, or a calibration check vapor, is unacceptable. 

      The operator of the breath test machine must also be sufficiently trained to administer the test with a full understanding of the issues that may compromise the test result, including the theory of operation, detailed procedure of operation, and practical experience. Specifically in California, the operator is required to take 4 to 6 hours of training including: Theory of operation; detailed procedure of operation; practical experience; and precautionary checklist, The must then must pass a written and/or practical examination. State of California, Forensic Alcohol Analysis, Title 17, Reg. 76, No. 24, 6-12-76, section 1221.4(a)(3).  The NHTSA also has established some recommendations for Breath Alcohol Technicians. The Committee recommends instruction on: Alcohol and the human body (3 hrs); Operation of the machine, including a description of the method and a demonstration of the method (3 hrs); Legal aspects of the test and methods employed (5 hrs); Supplemental information, including nomenclature (3 hrs); Laboratory participation, including testing reference samples and actual subjects (10 hrs); and a formal examination (1 hr).  The minimum total training time required is 25 hours. Remote sampling by law enforcement requires the officer to receive at least 3 hours of instruction in: Principles and use of the device employed; Scientific significance of the test; and Proper techniques for the submission of the specimens.   National Safety Council Committee on Alcohol and Other Drugs, Committee Handbook, 1996, Appendix D, Recommendations of the Ad Hoc Committee on Testing and Training, 1968, pp. 90, 91.

The breath machines can be subject to RFI/electromagnetic interference that is a function of the intensity, frequency, and direction of the electromagnetic field and the characteristics of the electronic equipment in that field. The electromagnetic field contains energy which could induce voltages or currents within another electronic system in the field. [Department of Transportation, National Highway Safety Administration, DOT HS-806-400, May 1983.]  Many machines, especially those whose casing is made out of metal, are fairly well shielded against electromagnetic interference. However, even with this design, there are generally openings in the casing that allow an electromagnetic field to enter, creating EMI. The surface mounted devices in the newer instrumentation are not as susceptible as the older models to generating or receiving EMI. Even so, some shielding is still needed in specific configurations. The addition of communications ports (phone lines, high speed cable, usb ports) opens up new locations for EMI to enter. 

      Volatile chemicals present in the body can also affect the BrAC/BAC ratio. When the subject being tested has been recently or chronically exposed to these solvents, a quantity of the solvent may be present in the breath if they are taken into the body through inhalation of dermal absorption, circulated in the blood stream, and get stored in the body fat. These substances, like alcohol, are exhaled in the breath.  These voatile chemicals can also be found in alcohol-addicted subjects. Volatile organic substances partition from the blood into the alveolar air via the same mechanism as ethanol, oxygen, and carbon dioxide. Each chemical has its own temperature-dependent partition ratio.  However, breath machines will automatically apply a 2100:1 partition ratio to an absorbing substance, regardless of the type and origin. 

      This lack of specificity in identifying ethanol is one of the major issues concerning the reliability of breath testing. Modern machines do not positively identify ethanol. Depending on the technology employed, the machine identifies a chemical absorbing at one or more wavelengths, and/or identifies a chemical capable of being oxidized in the electrochemical cell. All machines have differing capabilities of detecting these substances, and not all machines alert the operator that an interfering substance is being misidentified as ethanol. Since breath testing machines do not specifically identify ethanol only, select substances found in the breath may contaminate the true breath ethanol and cause a false high reading 

      Some specific volatile chemicals have been found to affect the partition ratio.  Acetone is a chemical naturally found in small quantities in the human breath, and in larger quantities under certain circumstances. Exposure to acetone can result in the detection of acetone on the breath for several hours. The occurrence of acetone in a subject’s breath is of great concern in the field of breath alcohol testing because acetone absorbs light in the 3.39 and 3.48 micron region. Since acetone has a greater absorption at 3.39 microns than 3.48 (opposite the absorption pattern of ethanol), the relationship between the two wavelengths will indicate if there is an interfering substance. 

      Ether is also a problem with IR-based machines that employ absorption in the 3.4 micron region because diethyl ether vapor interferes with breath alcohol analysis by instruments based on infrared absorption at 9.5 microns. The Draeger® Alcotest 7110, Siemens Alcomat® V5.2F, and Seres Ethylometre® 679T show substantial readings, without triggering the interferant flag. Other chemicals on the breath may affect select machines more than others. Unfortunately, not all machines have been tested with all substances to see if the machine will register a false-positive or be able to detect that an interferant is present. Specifically Toluene and isomeric xylene (ortho, meta, and para) can cause false ethanol results on many machines that use infrared spectrophotometry.

In some jurisdictions, statutes have dictated that the 2100:1 conversion ratio be applied to all persons. However, breath alcohol testing machines in the United States already automatically use 2100:1 to convert the breath to the equivalent blood result. However, variances occur. There are many known and accepted factors contributing to variances between BAC and BrAC,  BAC is affected by analytical error, pharmacokinetics, and blood source. BrAC is affected by Breath temperature, the volume of air expired, analytical error, other substances on the breath, and mouth alcohol.  The partition coefficient ratio is not the same in all people, and the standardized 2100-1 conversion ratio becomes subject to challenges because individual ratios can fluctuate, and it is well known that many errors in breath testing can be traced back to the variation in blood to breath ratios between subjects.

Medical issues and conditions can affect the final result on a breath testing machine. Diseases of the lung known as Chronic Obstructive Pulmonary Diseases and Chronic Obstructive Pulmonary Diseases, along with digestive system diseases such as Gastroesophageal Reflux Disease generally artificially increase the ethanol result by allowing alcohol to contaminate the mouth, or narrowing the bronchial passages such that the breathing pattern is altered. Persons with diabetes can see increases the concentration of interfering compounds in the breath.   

      In the absorptive state, the breath alcohol concentration reads higher than a simultaneous venous concentration due to the arteries receive the alcohol first, then the veins during absorption. The BrAC will be higher than the BAC in the lungs during absorption because it receives the arterial blood.  Blood differences can affect the BAC as well. A person with a high hematocrit will have a slightly higher concentration in the water fraction of whole blood, and in the air phase.   Also, aAlcohol contamination in the mouth can either be by the recent consumption of alcohol, regurgitation, vomiting, or a burp. Alcohol in the mouth takes time to dissipate, such that there is negligible effect on a breath test result. 

      The lungs vital capacity, or the difference between total capacity and normal breathing capacity. Some breath-alcohol analyzers will record a volume for the breath sample analyzed, but the volume measurements can be flawed in several ways.   The volume measurement feature only starts logging the volume when the flow rate is at or above the threshold rate. Also, is if the thermistor used to determine the flow and volume is out of calibration that can routinely show extremely large and falsely elevated volumes for breath samples are easy to identify. Machines that are reading the volume on the low side are the hardest to identify and have the most serious repercussions for the accused taking a test.

The original studies of human breath were conducted by French chemist Antoine Lavoisier between 1774 and 1785. Aside from defining respiration as the uptake of O2 and the output of CO2, Lavoisier’s invention, the “gasometer,” was the first instrument to make accurate measurements of the respiration gases. In the 1850s, physician John Hutchinson modified Lavoisier’s gasometer to make the first spirometer for measuring the volume of a patient’s breath. In 1874, British physician Francis Anstie, trapped the human breath and applied colorimetric analysis to study the fate of alcohol in the body. Building on this knowledge, actual analytical analysis of expired breath for blood alcohol concentration was first proposed in the 1920’s by Bogen. 

      Early versions of quantitative breath testing machines included the Drunkometer® (1938), Breathalyzer® (1954), and Alkotest® tube (1954). These early methods employed wet chemistry; generally, oxidation with potassium permanganate, potassium dichromate, or iodine pentoxide. Gas chromatography was employed in the early 1970’s, and, in 1974, infrared (IR) technology took the lead in popularity.


      In 1974, the National Highway Traffic Safety Administration (NHTSA) published the “Standards for Devices to Measure Breath Alcohol,” and listed devices which met the Federally mandated criteria.  State programs applying for Highway Safety Funding were restricted to using equipment listed as meeting the Federal criteria. [NHTSA (1974), 39, Fed.Reg 41399.  This  publication morphed into what is now called “Model Specifications for Evidential Breath Testing Devices,” including a list of conforming breath test machines on a “Conforming Products list.” What this meant is that State programs were no longer restricted to using devices listed as meeting the Federal criteria, though most states continue to limit their consideration to equipment on the list. [NHTSA (1983) 49 Fed.Reg48864]. In 1984 he NHTSA modified their specifications, changed the levels at which the machines were evaluated (0.020%, 0.040%, 0.08% and 0.160% instead of 0.050%, 0.101% and 0.151%), added a test for acetone detection, and expanded their definition of “alcohol” to included low molecular weight alcohol such as methanol and isopropanol. [NHTSA (1993) 58 FR 48705.] The latest list includes several machines that have become compliant since 1993. [NHTSA (2002) 67 Fed.Reg. 6209].

      Although breath alcohol tests are an indirect methodology for determining BAC, it is more popular than the traditional direct measurement of blood because breath alcohol testing has the beneficial advantages over blood testing in that it has a lower cost to the user agency; is a more rapid procedure, giving immediate results; and is less invasive than drawing blood.

      The breathing process involves air entering and exiting the human body through the oral or nasal passages, traveling through the trachea and bronchi, the non-cartilaginous bronchioles and the respiratory bronchioles to the alveoli. The process of gas exchange can be broken down into three functions: Ventilation; Perfusion; and Diffusion. 

      During ventilation, fresh air is moved into the lungs and carbon dioxide is expelled. The total lung capacity of the human lung is about 5700 cm. However, the entire lung is not used during normal breathing. Only a small amount of 500 cm, is inhaled and exhaled with each breath when the body is at rest,. Even when a subject provides maximum exhalation, a small amount of air still remains in the lungs. This amount, approximately 1200 cm3 is called the residual volume. The difference between the two (called the vital capacity, or “VC”), about 4500 cm3, is the amount that actually moves in and out of the lungs during exertion. Ventilation parameters become important in the ability of a subject to deliver an adequate breath sample, since most breath testing measuring devices employ either a pressure sensor or a minimum volume and flow rate and require a prolonged exhalation.

      Perfusion is a process whereby the blood from the right ventricle comes in contact with the pulmonary capillary bed. Since the capillary bed is next to the alveolar sacs, diffusion of gases to and from the blood and sacs occur.  The amount of ethanol that partitions between the blood and the lungs generally follows the basic rules of Henry’s Law. Henry’s Law states that the amount of gas which dissolves in a liquid is dependent on the partial pressure of the gas in contact with the liquid and the solubility coefficient of the gas in that particular liquid.  Thus, at a given fixed temperature, a numerical relationship between the substance in the gas phase can be related to the substance in the liquid phase. In breath alcohol testing, this implies that ethanol in the blood can yield a predictable concentration in the breath. The ratio commonly used in breath testing machines is 2100:1, meaning 2100 milliliters of breath will contain the same amount of ethanol as 1 milliliter of blood. Breath is generally expressed as grams per 210 liters of breath (g/210), or converted to a blood alcohol equivalent.  However, the application of Henry’s Law to breath alcohol testing is not without flaw. When alveolar air is exhaled, there is a re-equilibration between the air and the mucous membranes of the upper respiratory tract, and a corresponding drop in temperature of the breath. In addition, there can be considerable intra- and inter-subject variability, depending on the circumstances, of the blood to breath partition ratio.

      There are a variety of breath alcohol testing instruments available to a user agency. Many of the evidential devices are listed in the Federal Register Conforming Products List of Evidential Breath Measurement Devices. [NHTSA ( 2006) 71 Fed.Reg. 37159.] The current model specifications for evidential testing devices were first published in 1993 [NHTSA (1993) 58 Fed.Reg. 48705]. Other devices include alcohol screening devices designed to give an approximation of the alcohol level, passive alcohol sensors designed to detect the presence of alcohol only, and ignition interlock devices designed to prevent a driver from starting a vehicle if a certain BAC is present. The model specifications for screening devices was first developed in the early 90s [NHTSA (1994) 59 Fed.Reg. 61923, corrected at 59 FR 65128] to address screening test devices for workplace testing. A Federal Register Conforming Products List of Screening Devices to Measure Bodily Fluids [NHTSA (2007) 72 Fed.Reg. 4559] is published every couple of years.  

      Breath testing technology, on modern instruments, is either via a Taguchi semiconductor, fuel cell, infrared detection, or a combination of infrared and fuel cell. All of these technologies have been in use in science for decades. However, it has only been in the last few decades that the technology has been adapted for breath alcohol testing in a law enforcement setting. There are two major categories of screening test devices: Passive. Passive devices merely measure the air space around the suspect without requiring the subject to deliver a breath sample directly into the device; and Active. Active devices require the subject to blow directly into the device, what are popular in the law enforcement field because they are portable, low cost, and can be used in the field with ease.   The disadvantage is the screening devices have no safeguard to ensure that the tested air is essentially alveolar in composition, resulting in an outcomes that are falsely elevated   

      The devices also detect the class of compounds called “alcohols”  rather than ethyl alcohol specifically, generating a reading on a fuel cell device.  This is aggravatged when small amounts of ethyl alcohol and other compounds are produced in the body as a by-product of normal body functioning even without alcohol consumption.  Some states now prohibit the refusal of a breath-alcohol screening test and/or allowing the use of the test results in trial if either an observation period was observed or if the test subject ultimately refuses the evidential test.

      Most law enforcement agencies use roadside tester with either EC or SC technology commonly known as Preliminary Alcohol Sensors (PAS), and are advantageous because they are compact handheld devices.  An electrochemical cell (also referred to as a fuel cell) continuously transforms the chemical energy of a fuel and an oxidating material into electrical energy. The purpose of using electrochemical cells is for energy conversion. Semiconductor or “SC” devices detect gas through an increase in electrical conductivity when reducing gases are adsorbed on the sensor’s surface. Oxygen is adsorbed on the SnO2 bead/crystal with a negative charge, resulting in a positive charge in a space charge layer that prevents the free flow of an electric current. In the presence of alcohol vapor, the surface density of the negatively charged oxygen decreases, and the resistance of the sensor is changed.  IR Technology or infrared radiation spectrophotometry uses chemical absorption of infrared energy at specific wavelengths. The chemical concentration of that breath can be determined by measuring the degree of absorption of infrared light by breath contained in a chamber. The typical IR-based breath machine will have an IR detector at the end of the chamber which converts the IR energy to electrical energy. In the breath testing procedure, the machine generally analyzes ambient air, to determine a baseline voltage output. When the subject’s alcohol-laden breath is measured, there is a resultant voltage decrease. This drop is measured, and converted to a numerical result. However, IR’s use of a limited number of wavelengths rather than the entire spectrum can lead to false positives. If ethanol molecules are in the breath chamber, not only can other substances absorb the light as well, but particles in the breath chamber can scatter the light and keep some light from reaching the detector, thus decreasing the amount of light that reaches the detector increases the apparent BAC reading. This is the so-called “Tyndall” effect.