1) Definitions

These leaks are usually referred to as emissions or releases. They are simply inadvertent discharges of either chlorine liquid or vapor into the surrounding atmosphere. There are both minor and major leaks. When liquid chlorine is involved, the leak is almost always major.

Minor leaks usually occur at the start-up of a new installation or right after maintenance or inspection procedures. They usually are the result of a gasket failure, valve packing adjustment, or equipment malfunction.

Major leaks are generally referred to as catastrophic leaks. They include such things as container rupture, guillotine break in a pipe line under pressure, broken flexible connection, fusible plug failure, and/or repair work while a system is under chlorine supply pressure. Not to be overlooked is the real possibility of spontaneous combustion of steel and chlorine if the temperature reaches 483^oF for some reason.

2) Frequency and Magnitude of Chlorine Leaks

A list of the chlorine accidents reported to the Chlorine Institute over the past 50 years indicates by frequency and magnitude of chlorine emission. The following category of causes is in order of potential hazard:

• Fire
• Flexible connection failure
• Fusible plug failure
• Freak accidents caused by carelessness and ignorance
• Valve packing failure
• Piping failure
• Equipment failure
• Container failure
• Chlorine pressure gages

(1) Due to Fire

Fire was listed by the Chlorine Institute as the most serious hazard primarily because companies that stored swimming pool chemicals were often wiped out when stored granular hypochlorite suffered a spontaneous explosion. If 150-1b cylinders of chlorine were in the area, these too would explode because the fusible plug melts at a few degrees after the cylinder becomes skin-full. Ton containers are different. When they become skin-full, the dished heads blow outward without rupturing the container. This maneuver increases the volume of the container which is sufficient to allow the necessary expansion of the chlorine during the next 4^oF rise required for the fusible plug to melt. Fortunately, such fires as described above do not happen in water or wastewater treatment plants. Moreover, the plants that use modern chlorination or sulfonation equipment handle these vapors under a vacuum so that excessive pressures, for whatever reason, rarely if ever occur.

In the early years of ton container usage, the industry suffered container explosions due to the spontaneous combustion of nitrogen trichloride (NCl₃). When chlorine is made from electrolytic cell water containing ammonia N in sufficient quantities, the chlorine produced will contain NCl3, which is soluble in liquid chlorine. When the liquid is withdrawn, the NCl₃ evolves as a vapor and will explode. This first occurred at a New York City water treatment plant in the 1920s. It became a simple matter to remove the ammonia N from the cell water, but Americans did not do a very good job of exporting this information. In 1981, there were such explosions at the plant of local chlor-alkali located Bogota, Colombia, and S.A. It was reported that more than a dozen U.S.-made ton containers with their dished heads still intact although they had been blown from a concave to a convex position.

There never has been any report of a container failure of 150-1b cylinders & one-ton containers at any potable water, wastewater, or industrial cooling water chlorination or sulfonation installation in the United States or Canada. Fires in these locations are practically unheard of. If a fire were to occur that caused a major liquid chlorine release it would be vaporized immediately and would disappear in the chimney effect of the fire.

(2) Due to Flexible Connection Failure

Probably the most frequent cause of chlorine emissions resulting in overexposure to personnel is the failure of connecting lines between the container and the metering and control equipment. These connecting lines are traditionally made of annealed copper, 2000 psi strength, and cadmium plated. Copper is used because it is flexible and has the proper structural strength; however, at each container change, the chlorine remaining inside the tubing will absorb moisture from the atmosphere, and a cycle of corrosion will begin. Therefore, any flexible connection has a life dependent upon how many times it is disconnected and left open to the atmosphere. To check the worthiness of a flexible connection, bend it carefully; if it screeches slightly, it is due for replacement. The noise produced by the bending is indicative of the magnitude of corrosion products on the inside of the tube.

(3) Due to Fusible Plugs Failure

Fusible plug failure without any evidence of elevated temperature caused by fire or direct sunlight is next in order of hazard magnitude. A fusible plug which is supposed to melt at 158? may leak from corrosion or a poor bond between the lead alloy and the plug retainer. There is only one fusible plug on a 150-1b cylinder, at the base of the outlet valve . There are three in each of the dished heads of a ton container. Gas emissions through faulty fusible plugs have been numerous enough to question their safety value.

(4) Due to Freak Accidents caused by Carelessness and Ignorance

Carelessness is high on the list of causes of chlorine accidents, sometimes involving unusual situations. One case involved a 6-inch buried chlorine gas line originally and entirely within the property of two chemical plants. Adjoining property became subdivided 20 years later, and some of the underground piping, including that for chlorine gas, became part of the pavement area of new streets. This chlorine gas line was first cut into by mistake by a welder who was supposedly inactivating obsolete natural gas lines. The heat from the torch burned a small hole in the pipe, so that the chlorine could support combustion of the carbon steel pipe. Within seconds, the hole was approximately 8 inches in diameter, resulting in an almost immediate discharge of the contents of 7000 feet of 6-inch pipe. A few months later, on successive days, this same line was broken twice again by a backhol excavating for additional underground lines. After these experiences, the decision was made to lower the pipe from its former depth of approximately 2 ft to approximately 5 ft.

(5) Due to Valve Packing Failure

Valve packing failures have never caused any serious problems. If the leak is minor, it can often be corrected by tightening the packing nut. Serious leaks are taken care of by the application of a proper container safety kit.

(6) Due to Piping Failure

Piping failures have been rare, and are sometimes the results of using improper materials. Lines carrying liquid chlorine can be a potential hazard. One of the few fatalities attributed to chlorine in recent years was caused by the failure of a liquid chlorine line. The worker's death was the result of his constricted breathing caused by the amount of chlorine fumes, which made it impossible for him to climb out of the confined location of the leak.

(7) Due to Equipment Failure

Equipment failure usually refers to vaporizers used between the containers and the chlorine metering and control equipment. These failures are due primarily to corrosion by the chlorine, and are a function of the amount of wall thickness versus the amount of chlorine passing through the vaporizer. The frequency of this type of accident is low.

(8) Due to Container Failure

Container failures, except those caused by fire, are rare. The chlorine packagers through-out the United States and Canada are keenly aware of the potential hazards of handling chlorine containers. This has resulted in a strict monitoring of container condition. Perusal of the Chlorine Institute accident reports indicates two container failures over a period of 15 years. Considering that shipments in 150-1b and ton containers amount to nearly 100,000 per year, this sort of accident is a rarity and is usually a result of using an over-age cylinder.

One of these accidents occurred at a military base. A ton container was being loaded at a dock. It slipped out of the sling and fell on its end. The tank split at the "chime," which is the joint between the container and the dished head. Investigation of this accident revealed that this container had been in service by the military for 45 years.

(9) Due to Chlorine Pressure Gauge

This type of gage is vulnerable to leaks because of the silver or hastalloy diaphragm that protects the brass Borden tube. The Borden tube flexes to indicate pressure changes. Constant flexing of the diaphragm causes metal fatigue that results in diaphragm failure. When this occurs, the brass Borden tube corrodes quickly from the inherent moisture in the chlorine. This results in a severe leak. These gages should be replaced about once every five years.

3) Calculating Chlorine Leak Rates

All of the following calculations are based upon the Chlorine Institute formula shown in its Pamphlet No. 74 (1982) under 2.4.1, "Release rate Formulas."

(1) Liquid Release

Where:
A = area of opening to atmosphere, ft²
P1 = upstream pressure, psi
P2 = downstream pressure, psi
r = density of liquid chlorine upstream from the opening to atmosphere, lb/ft³

(2) Vapor Release

Where:
A = area of opening to the atmosphere, ft²
P = upstream pressure, psi
V = upstream specific volume (1/r), ft /lb

To be continued. Please press the next button....