Here is what 3,500€ will buy you

Overview

In our proposed estimation method, three different benefits of online monitoring systems are considered:

1. Decreased Repair costs - troubleshooting faults early is often cheaper than waiting for a major failure. Online monitoring increases the chance that faults are detected at an early stage, decreasing the expected repair costs.
    
2. Decreased Loss of production - when major failures occur, the primary function of a transformer cannot be fulfilled anymore. Depending on the transformer application, this can result in contractual penalties or lost business. If evolving faults are detected early, longer unplanned outages can be replaced by shorter planned outages, thus reducing the loss of production.
    
3. Decreased cost of offline DGA analysis - if no online-DGA system is installed, offline DGA measurements should be performed regularly instead. With an installed online-DGA system, these offline measurements are conducted less often or not at all, hence saving money.

These benefits are estimated over the lifetime of the electronics and summed up. Finally, the costs of a monitoring system are substracted. This results in an estimate for the Return on Investment (ROI).

Repair costs

Our model here is that each year, a transformer has a (0.44%) yearly fault rate, and that faults can result in three possible outcomes:

• They can be detected early. In that case, the transformer can be saved, but (20%) of the purchase price is needed to repair it.
   
• Undetected Faults result in a major failure with a probability of (97.5%). This will make a replacement of the transformer necessary, incurring a cost of (100%) of the transformer purchase price.
   
• Undetected Faults result in a catastrophic failure with a probability of the remaining (2.5%). This will lead to additional follow-up costs like environmental clean-up or compensation for bodily injury, incurring a cost of (300%) of the transformer purchase price.

The probability that faults are detected early is assumed to be (30%) if no ECOSENSE® ACTIVE PART is installed. With ECOSENSE® ACTIVE PART, this probability is increased. ECOSENSE® ACTIVE PART alone detects 23.26% of all transformer faults, therefore the combined detection probability can be calculated by reducing the detection gap by 23.26%:

combined_probability_fault_detection = 
  probability_without_ECOSENSE 
  + (100% - probability_without_ECOSENSE) 
    • probability_detection_ECOSENSE

30% + (100% - 30%) • 23.26% = 46.3%

Now the probabilities of outcomes can be calculated:

For the cost calculation, this distribution is multiplied by the failure rate and by the respective consequence costs. This gives the yearly expected repair costs as percentage of the purchase price with and without ECOSENSE® ACTIVE PART:

So with ECOSENSE® ACTIVE PART, repair costs are on average reduced by 0.35% of the transformer price per year, dropping down to 0.29% per year for a transformer equipped with ECOSENSE® ACTIVE PART. The expected yearly benefit is the difference multiplied with the transformer purchase price:

yearly_benefit_repair_costs = 
  (yearly_percentage_repair_costs_without_ECOSENSE 
    - yearly_percentage_repair_costs_with_ECOSENSE) 
  • transformer_purchase_cost

(0.35% per year - 0.29% per year) • 800000 dollars 
  = 480 dollars per year

Loss of production costs

The idea here is that detected faults lead to planned outages, whereas undetected faults lead to unplanned outages (when the transformer finally fails). Planned outages are much shorter on average, since no down-time is involved while replacement parts are ordered and job assignments are planned.

The yearly probability of an unplanned outage being avoided can be calculated by multiplying the yearly fault rate (0.44%), the detection probability of ECOSENSE® ACTIVE PART and the probability gap of fault detection without ECOSENSE® ACTIVE PART (30%):

reduced_loss_of_production 
= detection_probability_dga 
  • failure_rate 
  • (100% - detection_rate_without_monitoring)

0.44% per year • 23.26% • (100% - 30%) 
  = 0.072% per year

The expected yearly benefit is the avoidance probability multiplied with the hourly cost of an outage and the reduction in outage duration between planned and unplanned outages:

yearly_benefit_production_loss 
  = reduced_loss_of_production 
    • hourly_cost_production_loss 
    • (hours_per_unplanned_outage 
       - hours_per_planned_outage)

0.072% per year 
  • 100000 dollars/hour 
  • (168 hours - 16 hours) 
= 10944 dollars per year

Offline DGA costs

If no online DGA system is installed, offline DGA measurements are usually done instead. With ECOSENSE® ACTIVE PART, these lab checks are either reduced or avoided. The reduction in offline DGA costs can be calculated as follows:

fit_offline_dga = 
(cost_oil_analysis + cost_service_deployment) 
  • (service_interval_without_online_dga 
     - service_interval_with_online_dga)

(250 dollars + 150 dollars) 
  • (1 per year - 0 per year) 
= 400 dollars per year

Tally

The total yearly benefit of ECOSENSE® ACTIVE PART is the sum of the results of the three last sections:

yearly_benefit_total 
= yearly_benefit_repair_costs 
  + yearly_benefit_production_loss 
  + yearly_benefit_offline_dga

480 dollars per year 
  + 10944 dollars per year 
  + 400 dollars per year 
= 11824 dollars per year

Now we have the total yearly benefit. Together with the expected lifetime of the electronics and the costs, we can calculate the return on investment over the complete lifetime:

roi = yearly_benefit_total
        • lifetime_electronics 
      - cost_dga 
      - cost_installation

11824 dollars per year • 15 years 
  - 4000 dollars 
  - 4000 dollars 
= 169360 dollars