Why and how much for fire protection of power plant?

Fire protection measures can be subdivided into life-safety and investment categories.

Life-Safety Measures

Life-safety measures are considered to be mandatory by  fire codes, building codes, or safety codes. As such, the codes mandate specific types of substation fire protection, with very little flexibility in their selection.

Investment Considerations

Investment-related  fire protection is provided to protect assets, conserve revenue, and help maintain service to customers. This type of fire protection is not commonly mandated by legislation but is driven by economic reasons such as asset losses, revenue losses, and the possible loss of customers.

The selection of investment-related  fire protection can be done based on company policies and standards, insurance engineering recommendations, industry practices, specific codes and standards (IEEE 977 and NFPA 850), or by risk-based economic analysis.

TABLE 1 – Probability of Fire for Various Transformer Voltages

Source: IEEE 979, Guide for Substation Fire Protection.

The risk-based economic analysis is the evaluation of the investment measures in relation to the probability of  fire, the potential losses due to  fire, and the cost of the  fire protection measures.

This analysis requires a reasonable database of the probability of fires for the different hazard areas or types, an assessment of the effectiveness of the proposed fire protection measures, an estimate of the  fire loss costs, and a fair degree of engineering judgment. The potential losses usually include the equipment loss as well as an assessment of the lost revenue due to the outage resulting from the loss of equipment.

One of the most common risk-based economic analysis types is a benefit/cost analysis. This analysis is calculated using the following equation:

Normally, this ratio should be greater than one and preferably greater than two. A benefit/cost ratio of two means that the benefit (avoided  fire loss costs) is twice the cost of the  fire protection. Therefore it is a good investment.

One of the greatest difficulties is to estimate the frequency of  fire for the specific hazards. Some companies have extensive fire loss histories and loss databases. These databases can be used to estimate specific fire frequencies, but the results may be poor due to the small statistical sample size based on the company’s records. There are a number of other databases and reports that are in the public domain that provide useful data (i.e., NFPA data shop, EPRI Fire Induced Vulnerability Evaluation Methodology, and IEEE 979 Transformer Fire Survey). Table 1 shows the estimated probability of fire from the IEEE 979 Transformer Fire Survey.

Once the potential financial loss due to a fire has been calculated, the designer should input costs and effectiveness of any proposed  fire protection measure into the benefit/cost equation and determine the B/C ratio. If the B/C ratio is less than one, provision of the fire protection measure is not an acceptable investment.

Example of a Risk-Based Economic Analysis

The following is a simplified example of an analysis:

• A substation has four 138-kV single-phase oil-insulated transformers. One of these transformers is a spare and is located remote from the others. The load supplied by these transformers is 25 MW. A water-spray deluge system is being considered to suppress or control a  fire in the trans-formers. The deluge system is expected to protect the adjacent transformers, but not save the transformer that catches  fire.
  
  The estimated cost of a deluge system for all three transformers is $60,000. The individual transformers have a replacement value of $300,000.
• The utility’s chief financial officer questions whether this is a good investment.
• The company uses a discount rate of 10% and requires that all investments have a benefit/cost ratio greater than two. The assigned value of energy is $25/MW. The standard amortization period is 25 years.
• The annual frequency of  fire for a single 138-kV transformer is estimated as 0.00025  fires/year. Therefore, the combined frequency for the three transformers is 0.00075 fires/year.
• The estimated effectiveness of the deluge system protecting the adjacent transformers is 0.9. The deluge system will not save the transformer in which the fire originates; it is assumed to be a total loss.
• The  fire is assumed to originate in the center transformer in the bank of three single-phase transformers. It is assumed that in the absence of suppression, the fire will spread to destroy the two adjacent transformers. The spare transformer is not affected because it is remote from the other transformers.
• The estimated station outage period for this scenario is the difference between the outage time to replace all three transformers (a fire in the center transformer could destroy all three transformers) and the outage time to replace the center transformer (assuming the deluge system will protect the adjacent transformers).

The outage time to replace a single unit is  five days and to replace three units is 40 days. Therefore, the expected outage loss period is 35 days.

The expected lost revenue is 35 days × 24 h/day × 25 MW/h × $25/MW = $525,000.

The estimated annual revenue and equipment loss costs = (composite annual  fire frequency)  ×
(revenue loss for the station outage period + replacement value of the adjacent transformers) =
(0.00075 fires/year) × [$525,000 + (2 × $300,000)] = $843.75/year.

The net present value of the annual revenue and equipment losses for the 25-year amortization period at a discount rate of 10% = $7659.

The benefit/cost ratio = $7659/[$60,000 × (1.0/0.9)] = 0.115.

SOURCE // Substation Fire Protection – Don Delcourt, BC Hydro