The first point to make is that long interruptions by all major definitions are treated completely separately; that is, outside the context of PQ as defined here. This is because long interruptions are currently referred to as continuity of supply and not as poor PQ. The main reason for this is that the causes of lack of continuity of power supply mainly originate from the energy sector, i.e. power generation, transmission and distribution. What oftenhappens is that more mitigation of continuity of supply very often results in an increase in voltage dips, which, as will be seen later, generate some of the most troublesome poor PQ cost consequences. The reason for this is simple: to minimize the quantity and duration of these interruptions, utilities apply auto-reclosing techniques, which in turn result in increased number of voltage dips. The same thing happens when electric energy end users increase their power supply redundancy levels by connecting up to additional power lines. If no additional design measures are installed when doing this, the effect is usually an increase in the cost of voltage dips.

　　The costs of interruptions can be stated as the cost per event of a specific duration (most often 1 second, 1 minute, 1 hour, 4 hours, etc.). For comparisons and estimates these unit costs are usually specified by the following indices:

　　• The cost per peak power (based on maximum power demand).

　　• The cost per interrupted power (at the moment of the event).

　　• The cost per energy not supplied (level referred to the moment when an outage started or cost per peak energy not supplied when the cost is divided by the energy consumption at yearly peak demand level).

　　• The cost per energy consumption annually.

　　• The combination of both in one formula, i.e. the relative power cost plus relative energy cost.

　　Some examples are:

　　• The IEEE 493 [7] standard gives averages for assessing the cost of a single interruption of 1 h duration:

　　– Industrial plants: 643 $/kW+911 $/kWh.

　　– Commercial facilities: 21.77 $/kWh not delivered.

　　– Office buildings containing computers: $25.07 per peak kWh not delivered.

　　• The CEIDS/EPRI study found that the average cost of a 1 h outage to industrial and digital economy companies was $7795. A 3 min outage cost $2107. A 1 s outage cost $1477. For continuous process manufacturing, the outage costs were $12 654, $18 476, $14 746 for a 1 s, 3 min and 1 h outage respectively. For fabrication and essential services, the outage costs were below the continuous process manufacturing costs. Total average annual cost in the USA for zero-voltage events was $23 318.

　　• The breakdown of outage cost components for three of these groups of companies is presented in Table 18.2.

　　Table 18.2 Outage cost components [9] (Reproduced from Report of Primen for Electric Power Research Institute, Consortium for Electric Infrastructure to Support a Digital Society)