(79b) Abnormal Events - Operations Roundtable

This session will begin with a 30-45 minute kickoff presentation on the following four abnormal events.

? Flare Systems

? Reactor Safety

? Brittle Failures

? DCS Instrumentation Failures

Once the presentations are complete, the audience will have the opportunity to go to one of four breakout rooms where a specific topic will be discussed. In the breakout session audience members will be encouraged to share ?war stories? and safety related observations. Shared information will be recorded for later distribution; however, as the purpose of the round table is to increase awareness of various safety threats and not mitigations, the participants name, company, and any shared information on mitigation solutions will not be recorded.

After the four breakout sessions are complete and a quick break, the audience will be invited back as a group to review the summarized results of all four breakout sessions. The purpose in sharing abnormal event information as a roundtable is to increase the awareness of such threats for those people responsible for performing Process and Safety Hazards Analyses within each company.

The following are examples of the subject matter that will be reviewed in the initial kickoff presentation:

Flare System Issues:

Low steam flow to flare ? During a minor plant upset and flaring event, operators notice that the steam control valves to the flare were open much more than normal to control the smoke from the flare tip. Data analysis found that the steam flow to the flare tip was operating at about 60% of original design. Investigation of the steam meters, control valves and pressure drop confirmed that the flow was accurate and there was a restriction in the steam flow. A small tear was found in the steam strainers at the base of the flare stack.

Solids in flare drum ? Following a plant upset, liquid had accumulated in a wet flare drum. While trying to pump the liquids back to the process, the pump strainer kept plugging with small hard particles. Analysis of the liquids found gasoline and caustic had collected in the flare drum. The solid particles were polystyrene and polybutadiene compounds. Before an environmentally acceptable means of removing the solids and liquids could be arranged, another plant upset occurred and completely filled the wet flare drum causing back pressure on the flare header system. The level was reduced enough to keep the plant operating by continuous cleaning of the pump strainers and switching of the two pumps.

Cold Brittle Fractures:

Cold Box failures without pre-warning ? Suddenly the Cold Box vessel pressure indicator spiked from normal operating pressure of 3" W.C. to full scale of the pressure transmitter indicating a core exchanger leak. Gas sampling confirmed the presence of hydrogen and olefins, as well as other hydrocarbons. Upon shutdown of the plant and an inspection of the Cold Box core exchangers and piping, five separate leaks were detected as nozzle weld failures on one of the core exchangers. This cold box was 16 years old and had experienced 40 thermal cycles. This incident happened 6 months after the last thermal cycle.

DCS Instrumentation Issues:

Lightning Strike and Instrumentation Failures ? During a thunderstorm, the heater structure was hit by lightning causing voltage surge through the instrumentation and DCS system. This caused damage to both instruments and DCS circuit boards, shutting down the plant. Approximately 20 transmitters had to be replaced in the heater structure. The DCS equipment had to be rebooted and several controller cards had to be replaced. After rebooting the DCS, the plant was restarted.

Reactor Upsets:

Regeneration Gas Contamination ? A back-end acetylene reactor was in the process of a routine regeneration. The case had been steamed to remove free hydrocarbons and air addition was completed to facilitate the carbon removal. The reactor was flowing methane off gas to remove the moisture from the catalyst, when the methane refrigeration compressor shutdown. The methane off gas that was being used to dry the acetylene reactor became contaminated with ethylene. Temperature alarms sounded and excursions were noted on the acetylene reactor.