If you had any doubts that the “Writers” of UL2034 consistently put their “Vested Interest”, [the U. S. Dollar], above “Common Sense” Health & Life Safety Issues; just read the following:

“The ‘DRIFT’ Characteristics of all three of the Commonly used Sensors Reveals very clearly, the Urgent need C O Alarms to be required to have “Monitored Sensors.”
The 1992, 1995, and 1997 (Effective Oct.’ 1998), … ALL three contained a statement indicating that “Every Primary Component that’s Failure could cause the Alarm/Detector to FAIL to Alarm, …  MUST be Monitored … EXCEPT the SENSOR!!!


The following information is made available by GRI, [Gas Research Institute], and George E. Kerr, President / Founder of C O – EXPERTS.


Let us talk Sensors and Common Sense !

Be Aware/Sensor Info Warning ! ! !

IF…. Your are involved in CONSIDERING writing a Carbon Monoxide Ordinance, By-Law, or any form of Mandatory Requirement for your City, County, State, Province, or Nation, … it is IMPERATIVE  that you consider the LONG TERM IMPACT that such action can CREATE for your Citizens, … and your First Responder Teams.

Therefore, there are a number of VERY IMPORTANT FACTS that you NEED to KNOW that have been Very Successfully and Intentionally HIDDEN from the Public, the CPSC, the Fire/Rescue Responders; as well as the HVAC and Utility Company Investigators.
Let’s Talk Sensors, … and COMMON SENSE!!!
Like EVERYTHING in LIFE, there IS a “REASON” for EVERYTHING; however, it is Often VERY DIFFICULT for many of us to understand THE REAL REASON WHY.





Up to October, 1998, … 99% of ALL CO “Detectors” in North America were either MOS (metal oxide or tin oxide, solid state, semi-conductor) Sensors, … or “Color-Metric”/Bio-Mimetic/Chemi-Optic type Sensors.  Herein Lies the Problems!!!

EVERY MOS Sensor Type CO Alarm (or Detector) EVER Produced WILL go into FULL FALSE ALARM at the “END of LIFE”, unless it has become contaminated and KILLED … EVERY ONE … EVERY TIME!!!

(A former Exec VP of Marketing of American Sensors actually “Bragged that this made their detector Fail-Safe”, in a Fax sent to Distributors in May of 1999).

This means that EVENTUALLY AT LEAST ONE CO Alarm/Detector RESPONSE to EVERY HOME/DWELLING that contains such a Product is GUARANTEED!!!

Unfortunately, this FINAL “End of Life” Response is often ‘Preceded’ by Several Previous “Nuisance” ALARMS, due to the “Super-Sensitivity” caused by Sensor Aging.  (At this point in Time, … these products WILL NOT Reset, or Silence, without removing the Power Source.)  In addition, these Sensors are GREATLY EFFECTED  by Changes in HUMIDITY.  FAR MORE SENSITIVITY in HIGH HUMIDITY, … FAR LESS SENSITIVITY IN LOW HUMIDITY CONDITIONS.

GRI, [Gas Research Institute], collected and documented a large volume of Reports on CO Detector / Alarm “Responses” to C O Alarm activations over the years 1994-99. Among the many items covered in these Reports was the “TYPE” of Sensor Technology used in the CO Alarms that resulted in the CO Responders Investigation.

The following Chart indicates Very Clearly, that as the “Average Age” of the CO Alarms in our Nations Homes Increases, … so does the “Percentage” of these “Un-wanted / Nuisance” Alarms that are using a MOS, [Metal Oxide Semi-conductor], type of Sensor.
The scope of this Problem will continue to Increase EVERY DAY until these products are REMOVED from the homes in our Nation.

NFPA 720, Guidelines on the Installation and Maintenance of C O Alarms says that C O Sensors are to be Calibration Certified, [Checked], every three years. Since this is NOT practical in the case of “Residential Single Station C O Alarms”, it is the recommendation of C O – Experts that ALL Residential Detection / Alarm Products should be REQUIRED to have an Automatic “Kill Switch” installed to indicate the    “End of Useful Life”, … WELL BEFORE this Excessive Nuisance Alarm age is reached.




EVERY Color-Metric/Bio-Mimetic/Chemi-Optic sensor type CO Alarm (or Detector), that has not become inoperative due to a Dead Battery, or some other cause of failure, WILL go into a FULL FALSE ALARM.  (At this point and Time, … These products WILL NOT Reset, or Silence, without removing the Power Source.

The HUMIDITY EFFECTS  are Very Similar to the MOS Sensors.
This Sensor Technology was Invented by the University of California, … the patent was acquired by Mark Goldstein, President of the Quantum Group, and shortly prior to the CO Detector UL-2034 “Introduction” on April 30, 1992, was Licensed EXCLUSIVELY to First Alert, for “Retail” Sales.  This “Exclusive” was later revoked, and today this Sensor is being used by several Companies.
Because of this “Exclusive” Patent arrangement from ‘Prior’ 1992, to ‘about’ 1995, … EVERY OTHER CO Detector SOLD in North America during this Period, …  HAD TO BE A MOS TYPE SENSOR PRODUCT!!

Also, because the MOS Sensors of that ERA required so much ‘Power’ to make the Sensors function Properly, they HAD to be 120v, ‘Plug-in’ Detectors.  This FACT meant that First Alert had the ONLY ‘BATTERY POWERED’ CO Detector on the Market.

Unfortunately, this fact, resulted in the INCREASE OF ‘Mis-Information’ on CO Alarm installations by virtually EVERY OTHER Manufacturer throughout the Industry.  The ‘Marketing Decision’ was made that since the MOS Products HAD to be ‘Plugged In’, and that meant EITHER a ‘Cord Hanging Down the Wall’ or the ‘Expense’ of Installing a ‘Recessed Clock Receptacle’, or a ‘Junction Box’ for a Direct ‘Hardwired’ Detector, … these ‘Factors’ resulted in a VERY BAD, and ‘DEADLY’ DECISION  that was ‘MADE’, ‘Advertised’ and ‘Vigorously’ Promoted.

With the ‘SUCCESSFUL REVERSAL of FACTS’ that the CO Alarm Manufacturers had seen occur in the IONIZATION Sensor (for Smoke Detectors), and the ‘Residential Sprinkler’ Industry, … and in my Opinion with the ‘Blessing and Participation” of NFPA, … they TOO said: “We can make up a ‘STORY’ that sounds ‘Half-Way’ feasible, and it will fly”.

So, comes the ‘Story’ that since the ‘Specific Gravity’ of Carbon Monoxide (98.6 at 68 degrees F) is about the SAME as ‘Normal’ Air, … it is OK to ‘Mount’ the CO Alarm down by the floor, and because it will ‘MIX’ … (Eventually), with the Air in the ROOM, sufficiently to cause the Alarm/Detector to activate before the CO reaches a ‘Fatal’ Concentration for the ‘Average’ Healthy person.
While I can accept this to be true, in many cases, unfortunately, it CREATES another EXTREMELY Dangerous and LIKELY  scenario that is as follows:

Mrs. Jones purchases a ‘Direct Plug-in’ CO Alarm, goes home to install it, and them has the Problem of Deciding WHERE to plug it in.  The process of thought goes something like this:  “No, I don’t want to put it there, because that is where I plug in the Vacuum Cleaner, … and not there, because that is where the Kids plug in their toys or record player, etc., … Where do we have a receptacle the is NOT being used, … Oh yea, there is one Behind the Couch, or Behind the Drapes,” … or … some other place WHERE IT DOES NOT BELONG for EARLY DETECTION, … and LIFE-SAFETY!!


One absolute FACT that never changes is:


Important GRI Testing Conclusions:


One of the Most Important and Revealing Charts in the GRI, Topical Report Issued in March, 2000, Titled: “Carbon Monoxide Response Survey Analyses: FINAL Supplement 1994-99, was a Chart depicting the Percentage of Very Low, or Zero, CO Concentrations actually found during CO Alarm / Detector Responses between October, 1998, and April, 1999.

I believe that this is a DIRECT reflection of the Increased percentage of CO Alarms utilizing MOS type Sensors, following the Chicago “Nightmare”, [which was generally blamed on “Color-Metric” Technology problems].
The combination of these Growing Numbers, as well as the “Aging” of the MOS Detectors sold in the 1992-98 years, and the fact that once the “Filters” in these Sensors get soiled, they react to many, many different gases, [thereby causing “Truly False” Alarms], these factors combined to result in the following HIGH Percentages of 0 to 9 ppm CO discovered during First Responder Investigations.

1. First Alert …………………………..   84.2 %

2. Nighthawk ………………………….   52.5 %

3. American Sensor ………………..    43.3 %

Maybe this is WHY that by 1997-98, the MAJORITY of the people surveyed said that when their CO Alarm Activated,….. THEY CALLED NO ONE….. to “Check-Out” the Potential Problem  ! ! !

The Complete GRI Conclusions found on page 57 follows:

“The volume of CO alarm activations investigated by utilities participating in this project has fallen in each of the last two years.  The decrease in demand for investigations following activations is surely a welcome development for responders, reducing their costs as well as the strain on their resources.  However, the question of why this demand is diminishing is an important one that is as of yet unanswered.

Assumptions that this trend is due to an improvement in alarm performance are not supported by the data from this project since no increase was recorded in CO levels measured following activations (as would be expected if there were fewer nuisance activations).  With alarms activating at roughly the same CO levels measured in previous seasons the obvious question becomes “What is causing the drop in investigations volume?”  One possible scenario is that residents are not calling utilities for help following activations as often as then did in earlier years (Kramer and Tikalsky, 1999), found that a majority of July 1997 through June 1998 alarm activations in four North American Cities were not reported to responders).  If this is indeed happening, it would call into question the contribution alarms are making to resident safety.
The new gradually emerging CO detection technologies (i.e., electrochemical and infra-red) may show different activation characteristics.  These technologies have yet to see the kind of widespread exposure to residential conditions experienced by the two more-commonly available technologies.”




In 1995, a THIRD SENSOR TECHNOLOGY was introduced into the Residential CO Alarm Market by George E. Kerr & AIM Safety. The ELECTROCHEMICAL SENSOR has been Successfully used for Industrial/Commercial Gas Detection/Investigation for 20 years or more.
This Technology was considered far too expensive to be used in the Residential Market back in 1991 – 92, when Earlier Versions of C O Alarms/Detectors were being Developed.
While a Good Electrochemical Sensor is FAR MORE ACCURATE, RELIABLE and REPEATABLE than other Sensors now available, … Contrary to the two previously discussed Technologies, this type of Sensor will Decrease in Sensitivity to C O as it ages, … and eventually will FAIL to DETECT C O at a Level necessary to Guarantee Life-Safety.  Unfortunately, Most C O Alarms WILL NOT go into a ‘Fault or Error’ Warning when the Sensor has Drifted BELOW SAFE LEVELS!!!

(Extensive testing prior to the 1999  Recall of a Million Kidde-Nighthawk and LifeSaver CO Alarms / Detectors revealed that NONE, ZERO, had provided a ‘Warning’ that the Sensor had Failed Totally !

The ‘DRIFT’ Characteristics of all three of the Commonly used Sensors Reveals very clearly, the Urgent need C O Alarms to be required to have “Monitored Sensors.”
The 1992, 1995, and 1997 (Effective Oct.’ 1998), … ALL three contained a statement indicating that “Every Primary Component that’s Failure could cause the Alarm/Detector to FAIL to Alarm, …  MUST be Monitored … EXCEPT the SENSOR!!!

***** It is my personal and professional recommendation is that any & all C O Alarms mandated by Building Codes or By-Laws must specify that only C O Alarms, Detectors or Monitors must provide protection from Chronic Low Level C O Poisoning by sounding a minimum 85db Audible Alarm at 10PPM or less, must monitor Sensor activity; as well as having an automatic “Kill Switches” at or prior to 7.5 years from “Date of manufacture be considered accepted as in full compliance with required specifications.


George E. Kerr, President / Founder
CO – Experts, Div. of G. E. Kerr Companies, Inc.
More home carbon monoxide alarm sensor information, click here.


Back to CO-Experts Main Page


CO Experts

19299 Katrina Lane; Eldridge Missouri 65463-9102

Tel: 1-888-443-5377 Fax: 1-888-436-5377