Evaluating Losses in Electrical Equipment Focus on Transmission Utilities CNY Engineering Expo 2016 Syracuse, New York Arthur C. Depoian P.E.
Contents Introduction Present Value of ongoing energy costs Operating Conditions Assigning weighting to various conditions Methods to reduce losses over the long term Impediments November 2016 CNY Engineering Expo 2
INTRODUCTION November 2016 CNY Engineering Expo 3
Why? Simple Example! Loss evaluation inadvertently omitted from specification for 27 MVA transformer: Same specification, same manufacturer with loss evaluation: No-load losses reduced almost 6%! Total losses reduced almost 31%! Lack of loss evaluation or inadequate evaluation, while it may lower capital cost, results in significantly higher losses! November 2016 CNY Engineering Expo 4
How to chose? November 2016 CNY Engineering Expo 5
FORD CHEVY It looks like the Chevy will cost $150 more per year. But look at the conditions. What if you are different? How to Compare? Does the same question apply to commercial, industrial, and utility projects? November 2016 CNY Engineering Expo 6
$ NET PRESENT VALUE November 2016 CNY Engineering Expo 7
November 2016 CNY Engineering Expo 8
November 2016 CNY Engineering Expo 9
We calculate NPV using Excel or other software November 2016 CNY Engineering Expo 10
Excel Example A Interest Rate 5% B Cost of Electricity $0.10 per kwh C Annual operating hours 8760 Hours/yr D Annual Cost (B*C) $876.00 per kw of load E Annual Cost Escalation 3% per year Inputted information Year Annual Cost 1 $876 2 $902 3 $929 4 $957 5 $986 6 $1,016 7 $1,046 8 $1,077 9 $1,110 10 $1,143 11 $1,177 12 $1,213 13 $1,249 14 $1,286 15 $1,325 16 $1,365 17 $1,406 18 $1,448 19 $1,491 20 $1,536 In this example the NPV of a continuous 1.0 kw load or loss equals ~ $14,000 Modeled for 20 year period =NPV(C1,C8:C27) F Net Present Value $13,985 November 2016 CNY Engineering Expo 11
1 $876 2 $902 However, 3 changing $929 any assumptions can have a 4 $957 5 $986dramatic effect! 6 $1,016 7 $1,046 8 $1,077 9 $1,110 10 $1,143 11 $1,177 12 $1,213 13 $1,249 14 $1,286 15 $1,325 16 $1,365 17 $1,406 18 $1,448 19 $1,491 20 $1,536 21 $1,582 22 $1,630 23 $1,679 24 $1,729 25 $1,781 26 $1,834 27 $1,889 28 $1,946 29 $2,004 30 $2,064 F Net Present Value $19,201 Changing the evaluation period from 20 to 30 years has increased the NPV from ~ $14,000 to ~ $19,000 Modeled for 30 year period November 2016 CNY Engineering Expo 12
Changing Assumptions Interest Rate: A higher value yields a lower NPV. Time Horizon: A longer time yields a higher NPV. It is always best to obtain input from the owner or client business and finance management. Business that have a tight cash-flow likely have high interest rates and shorter horizons. Healthy non-profits and governments often have longer time horizons. Utilities may need to use guidelines provided by regulators. November 2016 CNY Engineering Expo 13
Loss Evaluation Value On the previous slides, we calculated the Net Present Value of 1.0 kw of Energy usage or of losses. This is the premium that the owner would be expected to pay for a system with 1.0 kw lower losses. It is typically expressed as Dollars/kW, such as $5,000/kW. This value may be weighted or reduced based upon various operating conditions. Some specifications only express the loss evaluation using the energy cost, Dollars/kWh, such as $0.05/kWh. However, without knowing other inputs per the previous slides, the bidder cannot calculate Dollars/kW and could not optimize his offer for the loss evaluation. November 2016 CNY Engineering Expo 14
OPERATING CONDITIONS November 2016 CNY Engineering Expo 15
How do different operating conditions affect our calculations? The previous slides showed the example of a 1.0 kw load operating continuously. But operating conditions vary greatly! A residential distribution transformer is always energized. No-load losses are present 100% of the time. A residential air-conditioner in Northern climates operates a limited numbers of hours per year. November 2016 CNY Engineering Expo 16
Varying operating conditions require more inputs. This distribution transformer installed in Northern climates operates most of its life at very low load. Low load losses are most important. This step-up transformer installed at a base load power plant operates most of its life are high load, so losses at that operating point are most important. November 2016 CNY Engineering Expo 17
What is your load profile? Equipment loads may vary throughout the day, the week, and the year. How will your equipment operate? Is it always on? Does is have significant losses at low or partial load? Can you describe the various loads in terms of hours per year at various operating points? Can you approximate operation at just a few points to simplify? November 2016 CNY Engineering Expo 18
WEIGHTING VARIOUS OPERATING CONDITIONS November 2016 CNY Engineering Expo 19
Variable Shunt Reactive Power Equipment Transmission Line Transmission Line Transmission Line Static Var Compensator STATCOM Synchronous Condenser All of these devices supply vars to or absorb vars from the transmission line for purposes of voltage regulation and other benefits not covered in this presentation. All are infinitely variable from maximum inductive to maximum capacitive ratings. The losses of all these devices vary with their outputs and possibly also with specific operating conditions such as ambient temperature and transmission line voltage. Design choices and equipment selections may significantly affect losses. But if the effort to evaluate losses is inadequate, the design and equipment will naturally shift towards the lowest initial cost, which may be a less than optimal solution. The speaker has extensive synchronous condenser experience, so the discussion will focus only on that technology. However, similar methodology applies to other equipment. November 2016 CNY Engineering Expo 20
Utility Synchronous Condenser Specifications Examples Utility # 1 Utility # 2 During the bid phase, when asked about loss evaluation, the utility provided a sample transformer specification and asked bidders to follow that. Operating conditions were well defined, but selected the maximum design ambient temperature, which I question. The system does not operate there! Dollars/kw were stated. Utility # 4 Utility # 3 When asked about loss evaluation, the response was, We do not have a method to evaluate synchronous condenser losses. Utility # 5 Operating conditions were well defined and this utility did select a typical average temperature and average conditions for evaluation. Operating conditions were well defined, but selected the maximum design ambient temperature and the lowest operating voltage, which I question. The system does not operate there! Dollars/kw were stated. Clearly, the methods to evaluate losses vary greatly. There is room for improvement. November 2016 CNY Engineering Expo 21
Possible Designs for Utility # 4 HV Bus Mvar Weight Each 6-pole machine Losses [kw] each Three 6-pole Machines One 2-pole Machine Design A One 2-pole machine Design B Total Losses [kw] Losses [kw] -90 5% 615 1845 1900 2238 0 70% 515 1545 1824 2100 120 20% 740 2220 2116 2453 225 5% 1131 3393 2921 2975 From real examples, but the data has been skewed to protect any proprietary information. Examining only this data and graph, it is not yet obvious which design yields the lowest overall losses. November 2016 CNY Engineering Expo 22
Include the weighting HV Bus Mvar Weight Each 6-pole machine Losses [kw] each Three 6-pole Machines One 2-pole Machine Design A One 2-pole machine Design B Total Losses [kw] Losses [kw] -90 5% 615 1845 1900 2238 0 70% 515 1545 1824 2100 120 20% 740 2220 2116 2453 225 5% 1131 3393 2921 2975 Weighted Total At $13,000/kW 1787 1941 2221 $ 23,236,200 $ 25,233,650 $ 28,876,250 Now it is obvious that the 1 st design has greater than a $5.5 M lower (better) loss evaluation than the 3 rd design! November 2016 CNY Engineering Expo 23
Weighting Summary It is necessary to at least estimate expected operating conditions so losses can be evaluated and measured at several representative points. In the previous example one solution had the clear advantage. However, different weighting may have changed the outcome. November 2016 CNY Engineering Expo 24
METHODS TO REDUCE LOSSES November 2016 CNY Engineering Expo 25
Don t oversize or over-specify As responsible engineers we desire that our designs and solutions fit the need for many years, usually many decades. A project that needs to be upgraded or replaced in only 5-10 years might be considered underengineered. We plan for the worst case long-term scenario. We need a size 6 today, but install a size 15 just in case we need it in 20 years. However! This likely means additional capital cost and even worse, additional losses, for a future need that may never materialize! November 2016 CNY Engineering Expo 26
What to consider Unless we are very sure of the growth, can we install a system that meets the known short-term needs? Can we make provisions for future expansion at the project site or even a different site? The likely advantage is lower capital and lower losses today and greater flexibility on the future expansion. November 2016 CNY Engineering Expo 27
Very large conference area, one switch. The lights in the room Same area, multiple switches. Even if the conference room is only partially occupied, all the lights are on. Only the needed lights are on. This is obvious, but the same principles are not always applied on large projects. November 2016 CNY Engineering Expo 28
Synchronous Condenser Example One system, total 225 Mvar capacity Three systems, 75 Mvar each, total 225 Mvar capacity OR R R R R C C C C However, most of the year, less the 50 Mvar capacity is adequate. More than 100 Mvar is needed only during summer peak or other contingencies, less than 5% of the time. This system experiences 100% of no-load losses all the time. Only one or two machines operate most of the year, so perhaps 1/3 or 2/3 of the no-load losses. November 2016 CNY Engineering Expo 29
HV Bus Mvar Same data as an earlier slide with switching Weight Each 6-pole machine Losses [kw] each Three 6-pole Machines One 2-pole Machine Design A One 2-pole machine Design B Three 6-pole Machines, switched Total Losses [kw] Losses [kw] Total On? -90 5% 615 1845 1900 2238 1845 3 0 70% 515 1545 1824 2100 515 1 120 20% 740 2220 2116 2453 1480 2 225 5% 1131 3393 2921 2975 3393 3 Weighted Total At $13,000/kW 1787 1941 2221 918 $ 23,236,200 $ 25,233,650 $ 28,876,250 $ 11,939,200 Move than $11 M evaluation difference! November 2016 CNY Engineering Expo 30
IMPEDIMENTS November 2016 CNY Engineering Expo 31
Why don t all evaluate losses? Large Shopping Mall Landlord Example Owner builds the store site, but occupant pays the utility costs Regulated Utility Example Utility only follows what the regulators ask or require. It is possible that the utility actually profits from the losses! If the purchaser or owner does not pay for losses, what is the incentive? November 2016 CNY Engineering Expo 32
Question # 1 Some products have standard energy use and cost labels, often mandated and/or designed by government authorities or industry groups, so these can always be trusted as a reliable estimate of your energy costs. True or False? False. Such labels may be reasonable estimates for assumed operating conditions. However, conditions and real operating costs may vary greatly by the application. Using only the label may lead to the wrong decision. November 2016 CNY Engineering Expo 33
Question # 2 If a higher interest rate is used in the NPV calculation, will the NPV be higher or lower? The NPV will be lower. A business that has a higher cost of money will not invest as much to reduce future expenses. November 2016 CNY Engineering Expo 34
Question # 3 Bidder A has offered a product that has 5% lower losses than Bidder B at design conditions. Bidder A also has a lower price. Should Bidder A win the contract? Trick question! There is not enough information. We must understand the losses at expected operating conditions, which may be quite different than design conditions. November 2016 CNY Engineering Expo 35
Question # 4 Name two possible methods to lower losses. 1. Do not over-design for the future worst case. Consider a system that is expandable in the future, if needed. 2. For systems with significant no-load losses, a modular approach may allow much of the losses to be completely turned-off for extended periods. November 2016 CNY Engineering Expo 36
Question # 5 Losses may not be evaluated because the owner or specifying engineer does not have adequate information, but now you know and will figure it out. Why might you still fail? The owner, who is spending the capital, is not paying for the losses. Sorry to say so, but the people controlling the money just don t care about the cost of losses. November 2016 CNY Engineering Expo 37
Questions or Discussion? Please consider attending my next presentation: Synchronous Condensers Applications on Electric Transmission Systems Capital District's 37th Celebration of NATIONAL ENGINEERS WEEK February 16 & 17, 2017 Albany Marriott, Wolf Road, Albany, NY For copy of presentation or additional information please contact: Arthur C. Depoian, P.E. GE Grid Solutions, LLC 518-583-1972 Arthur.Depoian@ge.com