“HF Boil-Off from an MHF Spill on Hot Highway Pavement” by Jim Eninger, Ph.D.

Jim EningerA major trucking spill of MHF onto hot highway pavement could vaporize thousands of pounds of hydrofluoric acid — more than the largest release of the 1986 Goldfish test.

Boil-off is one of several paths that highly toxic Hydrogen Fluoride (HF) can go airborne into the community. In the past, the primary focus has been on the release of superheated HF and Modified Hydrofluoric Acid (MHF) from oil refineries’ settler tanks, because flash atomization causes 100% to form a ground-hugging, highly toxic cloud (see: Flash Atomization of HF and MHF).

Mobil’s original intent to suppress flash atomization was to add sufficient additive that the MHF was subcooled, that is, the boiling point is above the operating temperature. However, even subcooled MHF spraying under high pressure from a rupture of a tank will break up into small droplets. While not nearly as fine as droplets from flash atomization, they nevertheless evaporate. Evaluation of how much was the objective of the large-scale MHF release tests conducted by Quest Consultants in 1993 (see: Superheated MHF Excluded from the Only Large-Scale Test Series).

Another significant path for MHF to go airborne is boil-off. This occurs, for example, if the rupture is in the top section of a settler tank. At the release of pressure, the superheated hydrocarbons (typically isobutane with a boiling point of 11F) and MHF (boiling point of  71F at 6-wt% sulfolane) are highly out of thermodynamic equilibrium at the tank process temperature of 106F. The tank’s contents will boil violently and much will be expelled from the tank before plunging in temperature to the low boiling point of the remaining liquid. The volatile hydrocarbons and MHF will then boil off at a rate governed by the transfer of heat from the environment through the tank walls to the liquid.

A second example of boil-off is the primary focus of this post as well as of item III from the post Five Points TRAA Science Advisory Panel Members Would Have Liked to Make at the Rule 1410 Refinery Committee Meeting.  Except for a 15 wt% sulfolane additive used for transportation, which increases the boiling point of HF modestly from 67F to 76F, MHF trucks have none of the other refinery-based mitigation systems. With an air temperature of 87F, highway surfaces can reach 143F. The stored solar heat in a hot highway will vaporize MHF.

Consider a massive spill from a truck transporting 33,000 lbs of 15 wt% MHF (boiling point 76F) onto a sun-baked highway pavement at 140F. The MHF immediately cools the surface of the pavement to its boiling point of 76F. The temperature response of the pavement with typical parameters is shown in Figure 1.

MHF Hot-Highway-Pavement Boiling-Off Transient Temperature Profiles of Pavement
Figure 1 – Temperature profiles in 11-inch-thick pavement initially at 140F with an 15 wt-% MHF spill (boiling point = 76F)

The classical mathematical transient heat-conduction solution predicts an initial infinite heat transfer to the MHF, where, in reality, a short period of film boiling will transition into the nucleate boiling regime. The rate of boiling falls off as temperature is conducted from deeper into the pavement, as shown in Figure 2 (blue curve). The cumulative amount of hydrofluoric acid vaporized is shown by the red curve.

MHF Boil-Off from Hot Highway Pavement - Rate & Cumulative Amount
Figure 2 – Hydrogen Fluoride boil-off rate and cumulative boil-off per square foot.

Consider, for example, a realistic hypothetical scenario where it takes emergency responders two hours to reach the site of a massive MHF spill and complete the hazardous task of neutralizing it. By that time, Figure 2 shows that 5 lbs of hydrofluoric acid would be vaporized by boil-off for every square foot of pavement. For a characteristic pavement area, consider a square with sides equal in length to four standard highway lane widths of 12 feet, which has a surface area of 2304 square feet (4 × 12 feet)². For this case, 11,520 lbs of hydrofluoric acid would be boiled off, an alarming amount that could be blown by a breeze into a roadside neighborhood. For comparison, this is 39% more HF than the 8,300 lbs released in the 1986 Goldfish test in the Nevada dessert.

And, this doesn’t include HF evaporation from the surface of the spill, which will happen even without an elevated pavement temperature.  This will be the subject of another blog post.

Notes about the Mathematical Model and the Solution

Two-dimensional transient heat conduction in a slab is one of the prime examples used in college-level teaching of partial differential equations. The curves presented in Figures 1 & 2 were calculated on an Excel spreadsheet from solutions presented in the lecture notes for the MIT undergraduate class “Linear Partial Differential Equations” taught by Dr. Matthew Hancock.

The Pavement parameters used were: