Final Program EXPRES 2012 - Conferences

generation are hard-pressed to operate at peak efficiencies.While insufficient efficiency of heat energy productionand distribution contributes to overall energy costs,efficiency analysis gains significance and puts into focusthe heart of energy generator - a boiler.The creation of relevant and comprehensiveperformance evidence for regional boiler plantspopulation in the region of the Province of Vojvodinaprovides numerous possibilities for implementing energymanagement approach and for critical evaluation ofaccepted practices in the field of energy efficiency. Acomprehensive insight and added awareness, whichpresented approach enables, in combination with expertand soft computing assistance can locate, recommend andspecify opportunities for increasing energy efficiency ofboiler houses and reduce irrational energy use and relatedcosts.A. Analysis approachSteam and hot water are used throughout industry,commerce and the public sector for a wide range ofprocesses and space heating requirements, and canrepresent significant proportion of the organization’senergy costs. It is, therefore, important for owners andoperators of boilers to ensure that the plant operates asefficient as possible and consequently reduce avoidableenergy costs.All sectors, industrial, commercial and public useefficiency as the basic performance indicator. However,defining, assessing and predicting energy efficiency is noteasy, primarily because it is affected by more factors thananticipated. Therefore, the basic intention of the analysisis to apply neural network approach which can createnecessary preconditions and guidelines for assessing andanticipating boiler operating performance.The approach is designed with a capability foridentifying, investigating and predicting possible influencefactors which should be taken into account for a reliableenergy efficiency analysis. The approach uses neuralnetwork computation which reveals opportunities forefficiency enhancement and makes good insight intoinfluence factors for efficiency.Also, the analysis alerts to the possibility of efficiencydrop in heat energy production, and timely preconditionsfor suggesting effective measures to ensure efficiency ofpower plants.B. Model formulationOperating performance analysis is based oncomputational model which uses data from energy surveysand measurements in the group of randomly selected 65boilers. The audit has been carried out at over 50 sitescovering the representative range of industrial, public andcommercial users of steam and hot water. The samplecovers approximately 25% of all boilers in the Province ofVojvodina which provides reliability and relevance ofobtained results.For this purpose, 65 × 5 experimental measurements arecollected (input parameters). About 300 of those are usedas training data for neural network model and the rest (25)is used for testing.Presented approach develops a model for assessing andpredicting efficiency based on real (measured) operatingperformance. The approach understands the use of neuralnetwork model for analyzing and identifying operationalscenarios which lead to peak efficiencies.The significance of applied method is in providingextensive possibilities for identification of inefficiencies.Proposed approach can be effectively used to assessoperating performance and find optimum operationalregime. The significance is more noticeable in predictingthe response of a complex energy system that cannot beeasily modeled mathematically.C. Auditing methodologyThe first step is selecting necessary parameters for eachsite and particular boiler plant. Chosen relevantinformation and parameters are grouped into boilernameplate information, boiler operational performance,occupation dynamic, capacity engagement, flue gascomposition and state or conditions of the plant and itssubsystems.Techniques used to carry out each audit are: metering,internal monitoring records analysis, engineeringcalculations and specifics based on opinions of operationalstaffs in energy department of each company [3]. Throughmeasurements, direct communications and interviews witha responsible person in the energy department of eachcompany or institution and physical audit of the plant,necessary information is collected and organized inseveral categories in the following way:Common information about the company: Activity;Sector.Common information for selected boiler: Basic /Optional fuel; Boiler type (steam or hot-water boiler);Exploitation time (years); Boiler purpose (technologyor heating); Capacity [t/h] and [MW]; Operationalparameters [oC/bar]; Plant engagement (permanent,seasonal, etc).Boiler operational performance: The most usual loadrange regarding to nominal load capacity [%]; Annualworking hours [h/a]; Specific fuel consumption[m3/h, kg/h].Measurement parameters: Flue gas temperature [oC];Flue gas velocity [m/s]; Load range during themeasuring of flue gas composition [%].Calculated performances: Flue gas flow rate [m³/h];Boiler efficiency according to GCV [%]; Boilerefficiency according to NCV [%]; Flue gas losses[%]; Annual fuel consumption [GWh/a]Combustion quality (flue gas composition): O2 [%] /λ [%]; CO2 [%]; SO2 [ppm]; CO [ppm]; NO [ppm];NOX [ppm] and H2 [ppm]; MCO [kg/h] and MNOX[kg/h].Monitoring and control existence: Stationary andcontinuity oxygen control; Measurement of heatenergy delivery rate; Flow rate measurement of boilerfeed water and condensate return; Combustion qualitycontrol, etc.Control type: Manual; Semiautomatic; Automatic.Reconstruction and/or revitalization record: Burnersand associated equipment; Heat recovery systems(economizer); Pipes; Automatic control equipmentetc.96