METERING AND MONITORING GUIDELINE FOR THE AGRI-PROCESSING SECTOR IN SOUTH AFRICA IN PARTNERSHIP WITH Schweizerische Eidgenossenschaft Confédération suisse Confederazione Svizzera Confederaziun svizra Swiss Confederation Federal Depar tment of Economic Affairs, Ed c at i o n a n d R es ea rc h EAER State Secretariat for Economic Affairs SECO Disclaimer Notice IFC does not guarantee the accuracy, reliability, or completeness of the content included in this work, or the conclusions or judgments described herein, and accepts no responsibility or liability for any omissions or errors (including, without limitation, typographical errors and technical errors) in the content whatsoever or for reliance thereon. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of the World Bank or the governments they represent. The contents of this work are intended for general informational purposes only and are not intended to constitute legal, securities, or investment advice, an opinion regarding the appropriateness of any investment, or a solicitation of any type. IFC, or its affliates, may have an investment in, provide other advice or services to, or otherwise have a financial interest in, certain of the companies and parties named herein. All other queries on rights and licenses, including subsidiary rights, should be addressed to IFC’s Corporate Relations Department, 2121 Pennsylvania Avenue, N.W., Washington, D.C. 20433. International Finance Corporation is an international organization established by Articles of Agreement among its member countries, and a member of the World Bank Group. All names, logos and trademarks of IFC are the property of IFC and you may not use any of such materials for any purpose without the express written consent of IFC. Additionally, “International Finance Corporation” and “IFC” are registered trademarks of IFC and are protected under international law. All other product names, trademarks and registered trademarks are property of their respective owners. Developed and published for distribution by the National Cleaner Production Centre, South Africa (NCPC-SA), and the International Finance Corporation (IFC) in South Africa ABOUT IFC IFC – a member of the World Bank Group – is the largest global development institution focused on the private sector in emerging markets. We work in more than 100 countries, using our capital, expertise, and influence to create markets and opportunities in developing countries. In fiscal year 2021, IFC committed a record $31.5 billion to private companies and financial institutions in developing countries, leveraging the power of the private sector to end extreme poverty and boost shared prosperity as economies grapple with the impacts of the COVID-19 pandemic. For more information, visit www.ifc.org. ABOUT THE NATIONAL CLEANER PRODUCTION CENTRE SOUTH AFRICA The National Cleaner Production Centre South Africa is a national support programme that drives the transition of South African industry towards a green economy through appropriate resource efficient and cleaner production (RECP) interventions. The NCPC-SA’s mission is to drive RECP in industrial and selected commercial and public sectors by equipping them to operate in an efficient, sustainable and competitive manner. Services and focus areas include industry and sector knowledge-sharing, company technical support; green skills development; and advocacy and awareness-raising. METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 2 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA AGRI-PROCESSING RESOURCE EFFICIENCY PROJECT IN SOUTH AFRICA The Metering and Monitoring Guideline for the Agri Processing Sector in South Africa was produced as part of a broader International Finance Corporation (IFC) Agri-Processing Resource Efficiency (APRE) project in South Africa, aimed to assist companies engaged in agricultural processing to transition to better water and resource efficiency practices. The Project is expected to help mitigate water supply risks in the sector, resulting from the water scarcity challenge in South Africa and throughout the region. The project is implemented in partnership with the Swiss State Secretariat for Economic Affairs (SECO). This report has been developed by the IFC and the NCPC-SA with inputs from the Danish Strategic Water Sector Cooperation (represented by the Royal Danish Embassy and Department of Water and Sanitation). water & sanitation Department: Water and Sanitation REPUBLIC OF SOUTH AFRICA ACKNOWLEDGEMENTS The development of this document was managed by Raymond Greig and Rong Chen (IFC). IFC commissioned Resource Innovations Africa (Pty) Ltd. We appreciate the effort of the key experts, Darrin McComb (Director, Resource Innovations Africa) and his team. The team is grateful to the World Bank Group colleagues for supporting the assessment and providing feedback on the report. We would like to thank Alfred Hartzenburg, Nonhlanhla Halimana and Robert Peck. We would also like to thank the NCPC-SA team: Victor Manavhela, Kevin Cillier, Lindani Ncwane and Julie Wells for the excellent production of the report. 3 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA TABLE OF CONTENTS 1. GLOSSARY 7 2. EXECUTIVE SUMMARY 9 2.1 Overview of Metering Implementation Approach 10 3. BACKGROUND 11 4. BUSINESS CASE FOR METERING 13 5. METERING PLANNING 17 5.1 Set Objectives 17 5.2 Determine Performance Metrics 19 5.3 Identify Measurement Points 19 5.3.1 Electrical Energy Metering 20 5.3.2 Water Metering 21 5.3.3 Boiler Systems Metering 22 5.3.4 Compressed air and Chiller Systems 22 5.4 Measurement Approach 24 5.5 Practical Example 25 5.5.1 Basic Measurement Plans 25 5.5.2 Standard Measurement Plans 26 5.5.3 Advance Measurement Plans 27 6. METERING TECHNOLOGIES AND STRATEGIES 29 6.1 Fluid Metering – Volumetric 30 6.1.1 Meter Selection 30 6.2 Fluid Metering - Qualitative 31 6.3 Boiler Efficiency and Steam Metering 34 6.3.1 Direct Steam Metering 34 6.3.2 Indirect Steam Metering 35 6.3.3 Indirect Condensate Return Metering 36 6.3.4 Generation Efficiency 37 6.4 Electrical Metering 38 6.4.1 Meter Selection 38 7. METERING COMMUNICATIONS AND STORAGE 41 7.1 Common mistakes made 42 7.2 Best practice for datasets 42 7.3 Case Study – Power Factor 43 4 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 8. METERING COSTS AND FINANCING OPTIONS 45 8.1 Water Metering 45 8.2 Electrical Energy Metering 46 8.3 Compressed Air Systems 46 8.4 Refrigeration / Chiller Systems 47 8.5 Boiler Systems 47 8.6 Overview Budget 48 9. DATA ANALYSES AND USAGE 50 9.1 Performance Measurement 50 9.2 Case Study 53 10. CASE STUDIES 57 10.1 Atlantis Water Supply Scheme 57 10.2 ABI Premier Place – Pheonix (McComb, 2016) 58 10.3 RFG Foods – Groot Drakenstein 59 11. REFERENCES 60 LIST OF FIGURES Figure 1: Overview of Metering Implementation Approach. 10 Figure 2: Comparison of cost escalation of electrical energy over the last 8 years. 13 Figure 3: Indication of both water and electricity cost escalation over the past 24years. 14 Figure 4: Setting Objectives and Targets. 17 Figure 5: Steps that feed into objectives and targets. 17 Figure 6: Relationship of objectives, targets and action plans. 18 Figure 7: Electrical Metering Schematic 20 Figure 8: Water Metering Schematic 21 Figure 9: Boiler Metering Schematic 22 Figure 10: Compressed Air Metering Schematic 23 Figure 11: Chiller Metering Schematic 23 Figure 12: Illustration of accuracy and precision. 29 Figure 13: SANS 1529 Accuracy requirements for meters over their specificied flow range. 29 Figure 14: Effluent Quality Monitoring Project. 34 Figure 15: Indirect method for determining steam production 36 Figure 16: Steam Generation Efficiency calculated using a mass balance approach. 36 Figure 17: Klein River Cheese steam system generation and system efficiency. 37 5 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA Figure 18: Logging data for the compressed air system including power, pressure and flow. 39 Figure 19: Overview of data collection approach. 41 Figure 20: Demand graph illustrating the difference between kVA supplied and kW drawn when the power factor equipment had tripped. 43 Figure 21: Monthly kWh data for a factory. 50 Figure 22: Annual rolling average of the same data. 50 Figure 23: Annual rolling average and cost. 51 Figure 24: Ratio of kwh to production. 51 Figure 25: Example of a scatter plot with linear regression through the data points. 52 Figure 26: Illustrates how to determine the plants baseload. 52 Figure 27: How to determine the R2 value. 53 Figure 28: X-Y plot with a linear regression. 54 Figure 29: Atlantis Recharge Quality and Level Sampling Record 57 Figure 30: Picture of the recovery and blending tanks. 58 Figure 31: RFG Electricity cost comparison July 2013 - June 2014 59 LIST OF TABLES Table 1: Overview of performance metrics. 19 Table 2: Overview of Metering Period Strategies 24 Table 3: Basic Measurement Plan - Expect a 2% saving on costs. 25 Table 4: Standard Measurement Plan - Expect a 5% saving on costs 26 Table 5: Advanced Measurement Plan - Expect a 10% saving on costs. 27 Table 6: An overview of the different water meter types and their characteristics. 31 Table 7: Example of municipality by-laws 32 Table 8: Extracts from City of Cape Town Waste Water By-law. 32 Table 9: Overview of water utilised and discharged over the course of a year. 34 Table 10: An overview of the different steam meter types. 35 Table 11: Overview of electrical energy meters and their characteristics. 39 Table 12: Budget price for the installation of a water meter. 45 Table 13: Budget price for the installation of quality meter types. 46 Table 14: Budget price for installing an electrical energy meter. 46 Table 15: Budget Cost for installing a compressed air system metering programme. 46 Table 16: Budget Cost for installing a chiller system metering programme. 47 Table 17: Budget Cost for installing a steam system metering programme. 47 Table 18: Budget Cost for installing a comprehensive metering programme. 48 Table 19: Overview of 12 month data for the plant. 53 Table 20: A plant’s 12 month data including regression analyses outputs. 55 6 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 1. GLOSSARY A (Amps) The measure of electricity flow in a conductor and usually measured with an ammeter or current transformer. COD Chemical oxygen demand (COD) is the amount of dissolved oxygen that must be present in water to oxidize chemical organic materials. COD is used to gauge the short-term impact wastewater effluents will have on the oxygen levels of receiving waters APRE Agri-Processing Resource Efficiency HDD Heating degree days are a measure of how much (in degrees), and for how long (in days), the outside air temperature was below a certain level. hl Hectolitre kl Kilolitre kg Kilogram kl Kilolitre kWh Kilowatt hour KPI Key Performance Indicator PPE Personal Protective Equipment M&V Measurement and Verification protocols NCPC-SA National Cleaner Production Center of South Africa R South African Rands SECO Swiss State Secretariat for Economic Affairs TDS Total Dissolved Solids WBG World Bank Group 7 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 8 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 2. EXECUTIVE SUMMARY South African companies in the Agri-Processing sector range between companies whose only metering data is derived from monthly utility bills (often estimates by the local council), to those companies that have hundreds of metering points in the plant measuring at one second intervals. In general, South Africa companies have limited sub metering systems in place and most rely on intensity targets (kWh or litre per kg production) to determine performance. With few exceptions, companies that have world class resource management systems have invested heavily in metering and measurement systems in order to drive efficiencies. These companies will typically target projects with a 2-3 year payback period and utilise the data from the metering systems to motivate for additional budget and approval of CAPEX. This guide is targeted at a technical or plant manager at an agri-processing facility and aims to provide clear steps and budgets to implement a measurement programme that is aligned to international standards. The sections include: 1. Metering Planning – which explains how to set objectives and targets, how to align the measurement approach accordingly and provides some practical examples for a typical agri-processing company. 2. Metering Technologies and Strategies – which reviews the different types of meters and their applications in the following areas: a. Water b. Water and effluent discharge quality c. Steam d. Electrical 3. Metering Communications and Storage – which reviews the common strategies to collecting data as well as ensuring that the database is properly compiled to allow for easy analyses. 4. Metering Costs and Financing Options – which provides enough information to compile a capital budget for the procurement of meters and an operational budget for the continued analyses of the data. 5. Data Analyses and Usage – which looks at common approaches to performance measurement and guidance on how to utilise statistical tools to better analyse patterns in data. Case studies and illustrations of how the information is applied are provided throughout the guide and makes reference to companies that the NCPC-SA has worked with over the years. The schematic on the following page summarises the proposed approach to implementing a metering programme. 9 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 2.1 Overview of Metering Implementation Approach Identifying the Need Section 2 and 3 Develop Metering Plan Set Objectives Section 5 Section 5.1 Determine kPI’s Section 5.2 REVIEW ANNUALLY ID Measuring Points Section 5.3 Select Meter and Data Storage Define Apporach Section 6 Section 5.4 Determine Budget Section 7 Install Meters Section 8 Analyses and Interpretation Section 9 Figure 1. Overview of Metering Implementation Approach. 10 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 3. BACKGROUND South Africa is a water-scarce country, and forecasts indicate that the gap between water supply and demand is expected to increase over the next 20 years. A World Bank 2030 Water Resources Group study projects a 17% supply:demand gap by 2030 which correlates with a prediction by the Institute of Security Studies. A significant contributor to this growth in demand is the industrial sector, of which agri-processing is an important component. In 2019, IFC and the State Secretariat for Economic Affairs of the Swiss Confederation (SECO), launched the Agri-Processing Resource Efficiency (APRE) programme to address this challenge. The programme aims to help the agri-processing sector in South Africa to improve sustainability and competitiveness, emphasising reductions in water use, along with related reductions in energy and fuel. Agri-processing includes sub-sectors such as animal slaughtering, dairy processing, fruit and vegetable processing, paper and pulp, sugar, brewing and malting and wineries. Several initiatives under APRE have identified energy and water metering and monitoring as a major opportunity for the agri-processing sector to reduce water and energy consumption. For example, several studies have shown that simple metering and monitoring of water inflow can help industries develop water use balances, identify leaks and water wasting procedures that can be fixed in the short term and ensure significant savings on the water consumption and related costs. However, many firms either do not see the full value of metering and monitoring practices or do not know how or where to start implementing metering and monitoring equipment / systems. This document has been specifically prepared for the agri-processing companies many of whom have very little sub- metering in their plants and still rely on manual inputs of billing data to analyse resource efficiency. To this extent, the guide focuses on what metering system should be in place, what those systems will cost to implement and how to go about analysing the data derived. The guide will not attempt to give detailed information on any one given topic but has referenced resources that have been developed for this purpose. The guide also does not attempt to bring formal Measurement and Verification (M&V) protocols standards required by many of the tax incentives but rather aligns with ISO 50006 approach to performance measurement. Companies looking to take advantage of the tax incentives will need to develop a monitoring programme that has the incentive requirements in mind. 11 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 12 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 4. BUSINESS CASE FOR METERING As South Africa’s population has grown, the resource and capacity requirements to sustain the growth have not kept up, resulting in supply constraints which inevitably impact on cost, availability and risk of interrupted supply. The electrical energy cost escalations and interrupted supply is a point in case. The chart below is the actual (blue bars) cost per kWh escalation over the past 8 years for a company in the agri-processing industry. Eskom’s published information1 for its industrial clients (brown bars) are also included for reference. $ 2,00 $ 2,00 $ 1,53 $ 1,43 $ 1,50 $ 1,33 $ 1,29 $ 1,21 $ 1,11 $ 1,04 $ 0,93 $ 1,00 $ 0,81 $ 0,70 $ 0,74 $ 0,63 $ 0,68 $ 0,52 $ 0,57 $ 0,50 $ 0,00 2014 2015 2016 2017 2018 2019 2020 2021 Agro Processing Producer Average Cost / kWh Eskom Average Industrial Sale Price / kWh electrical energy over the last 8 years. Figure 2. Comparison of cost escalation of There are a couple of points to note from the graph. 1. Cost Increase Rate Electrical energy cost increase has doubled over the past 7-8 years for the company and the trend is likely set to continue. 2. Cost Premium for being in a municipal boundary Companies purchasing from municipalities are paying a significant premium for being in a municipal boundary. While the average Eskom Industrial user has seen a R0.40 / kWh increase (78%) over the past 8 years the increase from those in municipal boundaries have seen a R0.96 / kWh increase (92%). 3. Re-structuring of tariff While traditionally the cost escalations related to increases passed on by Eskom, the increase realised by the company in 2021 (~30%) related to a change in the municipal tariff structure. This led to the company having to re-align its cost strategies from an electrical maximum demand based tariff to a time-of-use tariff. While the example cited above relates to electrical energy pricing, similar cost dynamics are being realised with water and effluent municipal services as supply and water treatment infrastructure capacity constraints are realised. Two of 1 https://www.eskom.co.za/CustomerCare/TariffsAndCharges/Pages/Tariff_History.aspx 13 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA South Africa’s largest Metropoles have faced water supply issues (Cape Town and Nelson Mandela Bay) and many smaller municipalities have had to revert to ground water to supplement surface water supplies. This is highlighted by the graph below which provides an overview of water and electricity cost escalation over the past 24 years as compared to inflation. South African electricity & water tariffs vs. inflation (CPO) Figure 3. Indication of both water and electricity cost escalation over the past 24 years.2 In order to mitigate against the cost escalations, agri processing companies will have to develop strategies that not only embrace efficiency but also take into account tariff billing patterns. These will be discussed in later sections of the guide. A detailed study of the effectiveness of advanced metering in companies was conducted by the Carbon Trust in the United Kingdom. They found in total a 5% METERING reduction in carbon footprint (proportional to energy consumption) of companies PAYBACK that adopted advanced metering systems3. Practically this means that investing 1-2 Investing 1-2 % % of the total utility budget to monitoring and metering should realise a payback of the total utility of 2 years. While it is important to note that installing meters won’t save anything, budget back into the information provided (as is demonstrated in many of the case studies in the metering and guide) will often be the basis for decisions that will realise savings without needing monitoring systems additional capital cost. These oppurtunities range from shifting to a more cost effective tariff, identifiying which compressor is operating more efficiently and should realise a increase subsequently increasing its utilisation, identifying leaks in the plant or being payback of 2 years. warned timeously when the plants power factor correction trips. 2 https://www.poweroptimal.com/the-price-of-water-and-electricity-in-south-africa-a-tale-of-two-tragedies/#chapter 3 The Carbon Trust (May 2007) Advanced metering for SMEs - Carbon and cost savings 14 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA Case Study – ABI Premier Place A resource optimisation assessment was conducted at ABI Premier Place in 2011 and the site was visited while no production was in place. The main meter was indicating a plant demand of 22.5 litres / minute which upon further investigation was attributable to an underground leak. The savings at the time was in excess of R85 000 / annum. Live logging or a downtime monitoring programme would have detected the leak far sooner. The plant has since put a concerted effort into its water efficiency programme which has seen its specific water usage (litres water per litre product) go from 5.5 in 2011 to 2.3 in 2013 and 1.8 in 2015. More details of this programme are provided in section 8.1 of this report. Legislative Compliance While metering programmes often result in significant financial savings, an additional benefit is realised through legal compliance. Companies require metering to be in place in order to obtain water permits and discharge licences which are often renewed on a 3 year cycle. The legislation includes groundwater and river water extraction. Many municipal permits require that a detailed balance be conducted for the submission with a minimisation programme in place to ensure that national LEGISLATIVE water efficiency objectives are being addressed. Similar legislation is being imposed COMPLIANCE in the energy sector with large users (> 180 000 GJ / year) currently required to While metering report on usage and very large users (>400 000 GJ / year) needing to develop programmes often energy management programmes. Some of the relevant sections of legislation are result in significant provided below. financial savings, an additional benefit • Government Notice 141 (2018) Instruction to install water metering devices is realised through • Government Notice 131 (2017) Taking water for irrigation purposes • National Water Act, 1998 (Act 36 of 1998) legal compliance. • National Energy Act, 2008 (Act No. 34 of 2008) • Regulations Regarding Registration Reporting on Energy Management 15 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 16 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 5. METERING PLANNING An effective metering and monitoring system will take years to implement and will require both equipment and human resources. An effective metering system design must avoid the following pitfalls: METERING PLANS • Gathering insufficient data to enable an accurate consumption analysis; Developing a • Gathering too much data that are never used for analysis; and metering plan • Gathering data in a format that cannot be easily used. should be the first step at any site Consideration should be given to both what needs to be measured as well as what will be done with the data once it is collected. To this extent, developing a metering that is initiating plan should be the first step at any site that is initiating a metering program. a metering programme. The steps for developing an effective metering plan are outlined below. What and how often Determine Define the Set Objectives performance measurment metrics points Where and how Figure 4. Setting Objectives and Targets. 5.1 Set Objectives In setting objectives one should consider key areas of resource usage, the variables impacting on the usage as well as the opportunities for improvement. 1 5 Organisation Existing Policy or OBJECTIVES opportunities Priorities database 2 4 Significant Prior reports Users of or audit Resources findings 3 Training Needs Figure 5: Steps that feed into objectives and targets. 17 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA It is important to differentiate between objectives, targets and action plans as at this point of the project cycle companies are setting long term goals which will underpin the proposed measurement plan. The figure below provides an overview of the key differences. OBJECTIVES TARGETS ACTION PLANS • Longer term • Specific • What? (maybe three • Measureable • Achievable • Who? years) • Relevant • When? • Specific • Timed • Consistent with • Is it complete? • Support the the policy objectives • Was it successful? Figure 6. Relationship of objectives, targets and action plans.4 As an example, an objective could be to reduce energy consumption by 20% within 5 years using 2020 as a baseline year. The associated targets would likely be: 1) Install live metering on all significant energy users by December 2021. 2) Replace all electrical heating elements with heat pumps by February 2022. 3) Increase chiller plant coefficient of performance (COP) (i.e., the cooling efficiency of the chiller) from 2.5 to 2.8 by March 2022.   4 United Nations Industrial Development Organization (2013) Practical Guide for Implementing an Energy Management System 18 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 5.2 Determine Performance Metrics Performance Indicators (PI’s) can be expressed by using a simple metric, ratio, or a model, depending on the nature of the activities being measured. Some examples of performance metrics are provided below: Table 1. Overview of performance metrics. Types Example Benefit Disadvantage Simple metric • Water consumption per • Readily available • Do not account for seasonal month • CFO’s historically interested variation or changes to • Annual Carbon footprint in total spend production volumes • Production downtime • Listed companies • Not able to benchmark required to report on total consumption Simple ratio • Energy per m2 • Readily available • Static factors not taken into • Water per tonne product • Easier to benchmark account • Rewards increase in production and not necessarily increase in efficiency Statistical analyses • A predictive equation based • Provides an accurate • Not easy to benchmark on historical performance indication of savings realised • Requires specific skill set to • Allows for developing work with effectively internal best practice • Statistical correlation not targets always found • Aligned to International Standards (ISO50006) There are usually factors that impact on resource utilisation performance. These typically include: 1. Static factors - quantifiable factor that significantly impacts performance and does not routinely change (i.e. facility size, product range, weekly shifts) 2. Relevant variables - quantifiable factor that significantly impacts performance and routinely changes (i.e. weather conductions, production output equipment load rates. 3. Interaction effect - when the effect of an independent variable on a dependent variable changes, depending on the value(s) of one or more other independent variables. When developing a metering plan one should identify both the performance metric and the factors influencing consumption as they in turn may need to be measured. In addition, the interaction of the factors influencing consumption on each other (interaction effect) should be considered. 5.3 Identify Measurement Points Drafting a simple schematic of the site can be a useful way of identifying key metering points. The schematics below provide an indication of the common metering points in an agri-processing plant as well as for advanced systems monitoring (i.e. Boiler, Compressed Air Plant and Chiller Plant). 19 5.3.1 Electrical Energy METERING Metering GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA AND MONITORING Common electrical metering points in a plant would include the main incomer, each major 5.3.1 as well Metering Electrical Energy department as each significant energy user (compressed air and chiller plants). Usually submetering Common electricalof intensive energy points metering in a processes plant wouldand include the main would compressors be incomer, advised. each major The main incomer department as well as each significant would energy usually user (compressed include air and the monitoring chiller plants). of quality Usually (specifically parameters submetering of energy power intensive factor) processes which and will be compressors would be advised. The main incomer would usually include the monitoring of quality parameters (specifically discussed in later sections. power factor) which will be discussed in later sections. Main Incomer M Q M M M Compressed Chiller plant Department Air Plant M M M M M M Process Process Compressor Compressor Chiller 1 Chiller 2 1 2 1 2 M = Metering Point Q = Quality Metering Point (power factor, harmonics, voltage imbalance) Figure 7: Electrical Figure 7: Electrical Metering SchematicMetering Schematic. 5.3.2 Water Metering Water metering systems in plants would usually include the main incoming points (municipal, ground, surface or rain water). Further sub metering would include the main departments and the main water consumption processes which would include the boiler, the cooling towers and the cleaning processes. Qualitative monitoring points would typically include weekly TDS, hardness and microbioligical coliforms tests from non-municipal water sources while water feeding into processes sensitive to water quality would have additional testing paramaters which may include chlorides, pH and the concentration of chemicals added to prevent corrosion or scaling in the boilers and cooling towers. 20 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 5.3.2 Water Metering Water metering systems in plants would usually include the main incoming points (municipal, ground, surface or rain water). Further sub metering would include the main departments and the main water consumption processes which would include the boiler, the cooling towers and the cleaning processes. Qualitative monitoring points would typically include weekly TDS, hardness and microbioligical coliforms tests from non-municipal water sources while water feeding P a g epH into processes sensitive to water quality would have additional testing paramaters which may include chlorides, 18 | and the concentration of chemicals added to prevent corrosion or scaling in the boilers and cooling towers. Q Main Water Feed M Alternative supply (I.E. Reclaim, ground or rain water) M M M M M Cleaning Department Boiler House Chiller plant Processes M M M M M M M M Q Q Q Q Q Process Process Feed Make- Cooling Cooling Manual CIP 1 2 Meter up Tower 1 Tower 2 Cleaning M Q M = Metering Point Q = Quality Metering Point (TDS, pH and COD) Figure 8: Water Metering Schematic. Figure 8: Water Metering Schematic 5.3.3 Boiler Systems Metering An overview of the metering points for boiler system would include water metering on the boiler make-up and feed with the associated quality parameter tests as discussed. The TDS of the boiler itself with automated TDS blow-down controller maintaining that parameter at the required set point. Boiler fuel usage is normally metered if it is a liquid fuel but solid fuels are harder to monitor continuously and will usually be based on billing data. A detailed discussion on boiler efficiency is included in section 5.3 as to how to implement a cost effective steam system monitoring program. 21 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 5.3.3 Boiler Systems Metering An overview of the metering points for boiler system would include water metering on the boiler make-up and feed with the associated quality parameter tests as discussed. The TDS of the boiler itself with automated TDS blow-down controller Pa maintaining that parameter at the required set point. Boiler fuel usage is normally metered if it is a liquid fuel ge but | 19 solid fuels are harder to monitor continuously and will usually be based on billing data. A detailed discussion on boiler efficiency is included in section 5.3 as to how to implement a cost effective steam system monitoring program. M Process 1 M Process Q TDS, Cl, Hardness 2 W Hot well M Make Up Water Q Stack O2 / CO2 / Temperature Feed Water W Boiler Blowdown F Q TDS W = Water Metering Point Q = Quality Metering Point F = Fuel Metering Point S = Steam Metering Point Figure Figure Boiler 9:9: Metering Boiler Schematic. Metering Schematic 5.3.4 Compressed Air and Chiller Systems 5.3.4 Compressed air and Chiller Systems An overview of the metering An overview of the points points for air for compressed metering and chiller systems compressed air andare provided chiller in the illustrations systems are provided below. Both in the entail live electrical metering of the compressors. Flow and pressure will usually be monitored after each air compressors illustrations below. Both entail live electrical metering of the compressors. Flow and pressure will and after the receiver as well as pressure at key points in the process. Water content in the compressed air will also be afterbe usually monitored themonitored after main receiver each in order toair compressors confirm and after the effectiveness theair of the receiver as well as pressure at key drying systems. points in the process. Water content in the compressed air will also be monitored after the main Chiller systems would monitor temperature set-points at key points in the process accurately and would also monitor receiver in order to confirm the effectiveness of the air drying systems. refrigerant pressure both before and after the compressor. Temperature and flow of the chilled water or product would be required to continuously gauge the cooling load requirement. These points will allow for the calculation of the COP Chiller (co-efficient of systems would performance) as monitor temperature well as the set-points system COP which at key inputs are important points in understanding into the process accurately and the oppurtunities would gains for efficiency also in the chiller monitor or refrigeration refrigerant systems. pressure Often both a first before step and in optimisation after assessments the compressor. is a calculation Temperature andon overall efficiencies based on first principals and system specifications. flow of the chilled water or product would be required to continuously gauge the cooling load requirement. These points will allow for the calculation of the COP (co-efficient of performance) as well as the system COP which are important inputs into understanding the oppurtunities for 22 systems. efficiency gains in the chiller or refrigeration Often a first step in optimisation assessments is a calculation on overall efficiencies based on first principals and system METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA e | 20 P a g Process F P E 1 Comp 1 P F Dryers Q F P Process E Comp 2 E P 1 Comp 1 F P Process F P F 2 Dryers Q E E = Electrical Metering Point Comp 2 F P Process P F Metering = Pressure P Point 2 E =F Electrical Metering Point = Flow Metering Point P =Q Pressure = Quality Metering Metering Point Point F = Flow Metering Point Figure 10: Compressed Air Metering Schematic Q = Quality Metering Point Figure 10:Compressed Air Metering Schematic. Figure 10: Compressed Air Metering Schematic Condenser P T W Condenser T E E P T Compressor 1 2 W T Expansion E E P T Valve Compressor 1 2 Receiver Expansion Direct X P T Valve Evaporator Receiver Direct X F Evaporator T T E = Electrical Metering Point T F T P = Pressure Metering Point E F Electrical Metering Point = Flow Metering Point = P = = Temperature T Pressure Metering Point Metering Point F = Water Metering = Metering WFlow Point Point T = Temperature Metering Point Figure 11: Chiller FigureMetering Schematic. 11: Chiller Metering Schematic   W = Water Metering Point Figure 11: Chiller Metering Schematic 23 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 5.4 Measurement Approach A plant’s measuring approach will need to balance capital costs for continuous logging systems and data requirements with potential savings impact. To this extent, there are four levels of resource metering each with their benefits and applications that they will be suited to. These are summarised in the table below. Table 2. Overview of Metering Period Strategies   One time / Spot Run time Short Term Long Term Measurements Measurements Measurements Monitoring Description Useful in determining Normally used where Equipment installed Term is typically more “baseline” activities hours of operation are for a limited period of than a year and and instantaneous the relevant variable time often the installation is resource use, permanent equipment performance, or loading Examples Boiler flue gas analyses Run time of fans or Fan or pump system Incoming electrical Lighting load pumps efficiency assessment and water supply Flow metering on Boiler steam meters cleaning applications Advantages Lowest capital cost Low capital cost Mid-level cost Highest accuracy Equipment usually Equipment usually Capable of assessing Data can be collected easy to use easy to use load variations automatically Non-intrusive Non-intrusive Relatively fast results Long term trends Fast results Useful for constant can be identified loads (seasonality and load Information can variance) be collected automatically Disadvantages Low accuracy Useful applications are Mid-level accuracy as High cost Limited application limited portable equipment is Most difficult to install Typically single Measures single utilised Installation process will operating parameter operating parameter Seasonal or take longer as CAPEX Information usually Requires additional occupancy variance expenditure approval manually processed calculations/ deficient is usually required. assumptions More difficult to install/ monitor 24 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 5.5 Practical Example The following tables provide templates for a typical agri-processing company with cooling and heating processes as well as water usage in the product itself. 5.5.1 Basic Measurement Plans Basic measurement plans are likely to achieve savings that require no further capital investment (choice or tariff, efficiency spot checks etc.). The table covers most of the main monitoring points and assumes that the data is manually entered into a local database for comparison purposes. The anticipated cost savings assume the data is incorporated into a resource optimisation plan and actively implemented as discussed previously. Table 3. Basic Measurement Plan - Expect a 2% saving on costs. Objective Utility Performance Static Relevant Frequency Measurement Location Indicator factor Variable Method Thermal Thermal Steam usage Steam Process Weekly Direct Boiler House Energy Cost Energy Pressure demand metering or Reduction Weather indirect conditions Thermal Thermal Boiler Boiler Process Weekly Direct Boiler House Energy Cost Energy efficiency Insulation demand metering and Reduction Losses Boiler Load indirect Stack Feedwater Temperature temperature Electrical Electrical Site electrical Ventilation Production Weekly Direct metering  Main Energy Cost energy energy Lighting Weather Incomer Reduction conditions Water Cost Water Site water Staff usage Production Weekly Direct metering  Main Reduction usage Leaks Weather Incomer conditions Effluent Cost Water Site water Daily Production Monthly Grab sample Effluent Reduction usage cleaning or composite Discharge Qualitative schedule sample Line parameters (COD, PH, TDS) 25 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 5.5.2 Standard Measurement Plans Standard measurement plans are also likely to achieve savings that require no or low levels of capital investment (leak determination and high level system enhancements etc.). The table covers most of the main monitoring points and assumes that the data is automativelly entered into a local or a hosted database for analyses purposes. A far higher degree of savings can be expected as more detailed data on an ongoing basis is avaiaible for the analyses. The standard measurement plan assumes that a competent resource is available to analyses the information and put the key observations forward for action by the departments in the plant. Table 4. Standard Measurement Plan - Expect a 5% saving on costs Objective Utility Performance Static Relevant Frequency Measurement Location Indicator factor Variable Method Thermal Thermal Steam usage Steam Process 30min Direct Boiler House Energy Cost Energy pressure demand metering or Reduction Weather indirect conditions Thermal Thermal Boiler Boiler Stack O2 30min Direct Boiler House Energy Cost Energy efficiency insulation levels metering and Reduction losses Stack indirect Stack temperature temperature Chiller Plant Electrical Chiller plant Circulation Production 30min Direct metering  Chiller plant Optimisation energy COP pumps and Weather cooling fans conditions Compressed Electrical kWh / m3 air Compressed Compressed 30min Direct metering  Compressed Air Plant energy produced air leaks air demand air plant Optimisation Water Cost Water Site water Staff usage Production Hourly Direct metering  Main Reduction usage Leaks Weather Incomer conditions Water Cost Water Department / Staff usage Production Hourly Direct metering  Different Reduction Process water Daily Weather Departments usage cleaning conditions / Processes schedule Water Cost Water Cooling Cooling Refrigeration Daily Direct metering  CIP Plant Reduction system water tower bleed load Feed usage rates Weather conditions Effluent Cost Water Site water Daily Production Weekly Grab sample Effluent Reduction usage cleaning or composite Discharge Qualitative schedule sample Line parameters (COD, PH, TDS) 26 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 5.5.3 Advance Measurement Plans Advanced measurement plans are likely to achieve significant savings that will likely require capital (installation of VSD’s, heat recovery systems etc.). External experts will be utilised to assist with conducting detailed system assessments and accompanying recommendations. The metering tools will be important in order to compile baselines which can be used to motivated for the capital required to implement the changes. Table 5. Advanced Measurement Plan - Expect a 10% saving on costs. Objective Utility Performance Static Relevant Frequency Measurement Location Indicator factor Variable Method Chiller Plant Electrical Chiller COP Circulation Production 30min Direct metering  Chiller Plant Optimisation energy and System pumps and Weather COP cooling fans conditions Compressed Electrical Compressor Compressed Compressed 1 second Direct metering  Compressed air plant energy kWh / m3 air leaks air demand Air Plant optimisation produced Fan system Electrical Electrical Friction Process air Annual Temporary Fan Location optimisation energy energy usage Losses demand metering Process pressure Pump Electrical Electrical  Motor Process Annual Temporary Pump system energy energy usage Losses demand metering Location optimisation Head Pressure Water Cost Water Cleaning Scheduled Production Daily Direct metering  CIP Plant Reduction water usage cleaning Feed Effluent Cost Water Site water Daily Production Hourly Direct metering  Effluent Reduction usage cleaning Discharge Qualitative schedule Line parameters (COD, PH, TDS) 27 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 28 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 6. METERING TECHNOLOGIES AND STRATEGIES When considering meter options it is important to understand key principles and how they apply to different metering technologies. Accurate, Repeatable Inaccurate, Repeatable Accurate, Non-repeatable Inaccurate, Non-repeatable Figure 12: Illustration of accuracy and precition. Accuracy – this is usually the first metric used to determine applicability of a meter to a particular system. No meter is 100-percent accurate and most manufacturers provide a range of accuracies in their product line and corresponding prices. Accuracy could be thought of as the difference between a measured value and that of the actual value. Published accuracies often will, and should, be referenced to specific calibration procedures including equipment-traceability according to SANS 1529 equipment and procedures. A meter’s accuracy will vary over the specified flow range of the meter. SANS 1529 requires that a meter’s reading inaccuracy is limited to ±5% at low flow and ±2% at high flow. Meters C and D in the figure below would not receive certification against the SANS 1529 standard. Meters should not only be selected according to required accuracy but also verified performance over the expected flow range. Lower zone Upper zone 5% d 2% a Relative error b 0% c -2% -5% qmin qt qp Flow rate qs Figure 13. SANS 1529 Accuracy requirements for meters over their specificied flow range. 29 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA Precision/Repeatability – the precision or repeatability of a measurement entails the ability to reproduce the same value (e.g., temperature, power, flow rate) with multiple measurements of the same parameter, under the same conditions. Turndown ratio – the turndown ratio refers to the flow rates over which a meter will maintain a certain accuracy and repeatability. For example, a steam flow meter that can measure accurately from 1 tonne/hr to 25 tonne/hr has a turndown ratio of 25:1. The larger the turndown ratio, the greater the range over which the meter can measure the parameter. Ease of installation – When making specific make-and-model decisions, it is important to understand any size and weight constraints, needs for specific diameters (or lengths) of straight pipe upstream and downstream of the meter, specific electrical and communications needs, and the overall environment the meter will operate in. Ongoing operations and maintenance – the lowest cost metering technology may not be the best choice if it has high associated maintenance costs (e.g., frequent service, recalibration, sensor replacement). As with most capital purchases, a life-cycle cost approach (including all capital and recurring costs) is recommended for decision making. Installation versus capital cost – in some situations, the cost to install a meter can be greater than the capital cost; this can be true where system shutdowns are necessary for meter installations, or where significant redesign efforts are needed to accommodate a meter’s physical size, weight, or required connection. In these cases, decision makers should consider alternative technologies that may have a high first cost but a much lower installed cost. A good example of this is the use of non-intrusive metering technologies (e.g. ultrasonic flow meters) that typically have a high capital cost but often a reduced installed cost. 6.1 Fluid Metering – Volumetric An exhaustive review of types of meters is beyond the scope of this guide and useful information in this regard has been published by the Department of Water Affairs (JE Van Zyl, 2011) . Important considerations for the selection of the meter are discussed below. 6.1.1 Meter Selection An overview of the different categories of water meters and the proposed application points are provided in the table below. The table compares water metering technologies with practical issues a plant may encounter. If the table is shaded red it indicates that it is less desirable and if green that indicates it is suitable given that specific parameter. The plants choice will depend on the specific application as well as the water utilised. For example, the “electrical” meter will monitor the amp load on the motor and the pressure on the pipeline and use the pump curve to determine the expected flow rate. This metering technique is suited to dirty water metering of large pipelines. It is not likely to be as accurate as some of the other meters however and would not be suited to billing applications. In certain circumstances high precision is required for product make-up purposes and so the capital cost of a Electromagnetic meter is warranted. The table also provides common points where the meters will be used for ease of reference. Most plants will utilise good quality water which make mechanical-turbine flow meters a cost effective choice for higher volume applications and mechanical-volumetric for the lower flow applications. 30 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA Table 6. An overview of the different water meter types and their characteristics. Type Mechanical – Mechanical Mechanical - Electro- Ultrasonic Electrical Volumetric - Jet Turbine magnetic Common pipe 15-40 15-40 40-500 300-2 000 400-4 000 Any pumped sizes (mm) application Sensitivity to Very Low Low - Medium High Medium High Low flow variation Sensitivity to High Medium Low Very Low Low Insensitive water quality Pressure loss High Low - Medium Medium Very Low Very Low Very Low Installation Low Low - Medium Medium High High Low limitations Electricity No No No Yes Yes Yes Installed Cost Low Low Medium High High Low Proposed Cooling Towers Cooling Towers Main line Process supply Not typically Effluent Application Hot well Admin block Borehole used Irrigation Admin block Ablutions Boiler Feed Borehole /River Ablutions 6.2 Fluid Metering - Qualitative Water qualitative parameters are important both for process water usage as well as effluent discharge. Key parameters and strategies to monitor these are discussed below. Incoming Water The incoming water quality should typically be analysed on an annual basis if supplied by a local municipality or on a monthly basis if the water is extracted from a natural water source or if the municipal supply quality varies. Typical tests of incoming water would include: • pH • Dissolved solids (TDS) • Alkalinity • Coliforms • Hardness levels BOILER BLOW Boiler and Cooling Towers DOWN Boilers should Weekly or monthly testing of water quality feed water would be conducted where be fitted with used as a make-up for the boiler and or evaporative cooling systems. Boiler systems TDS controllers to should be fitted with TDS controllers to ensure correct cycles of concentrations of control blow down the boiler at all times. Cooling tower and hot well make up systems are usually fitted and cycles of with pulse output meters that are linked to the chemical dosing systems, controlling concentration. scale and corrosion in these systems. 31 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA Effluent Very few companies continuously monitor their effluent discharge and both the council and the company rely on a single grab sample to determine the quality discharged. The discharge costs will often exceed the consumption costs as penalties are imposed for breaching the quality set-points prescribed by the local authorities. The table below provides an indication of the main discharge parameters tested for, but should a company be found to breach these, then additional tests and analyses would be undertaken and additional charges levied based on the defaulting parameter. The standard effluent charge calculation is usually a function of the quanitity of of water discharged, the organic load in the water discharged (COD) as well as a penalty for breaching the discharge parameters as quantified in the table below. Generally, if companies can show that their effluent discharge COD is consistently below 1000 mg / litre then the effluent quality cost parameter will not be levied. Similarly, municipal councils’ instead of installing metering on the effluent line will assume that 95% of the water supplied to the company is discharged to effluent unless metering can show otherwise. The actual calculations can generally be found in the by-laws published on the municpal web-site as indicated below. These tables would not be applicable to companies in rural settings discharging directly into waterways. Table 7: Example of municipality by-laws Municipality / Metropolitan Link to by-laws webiste Cape Town https://www.capetown.gov.za/Work%20and%20business/City-publications/policies- and-by-laws/policies-by-laws-and-publications Ekurhuleni https://www.ekurhuleni.gov.za/council/by-laws-policies/by-laws/ekurhuleni-by- laws-1.html Durban http://www.durban.gov.za/Resource_Centre/Pages/By-Laws.aspx The table below provides an indication of the qualitative thresholds municipalities adhere to when analysing effluent grab samples. Table 8. Extracts from City of Cape Town Waste Water By-law.5 Section a: General Not less than Not to exceed 1. Temperature at point of entry 0ºC 40ºC 2. Electrical conductivity at 25ºC 500 mS/m 3. pH Value at 25ºC 5.5 12.0 4. Chemical oxygen demand 5 000 mg/l 5 City of Cape Town Wastewater and Effluent By-Law, Promulgated 7 February 2014. 32 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA Section b: Chemical substances other than heavy metals – maximum concentrations Settleable solids (60 minutes) 50 mg/l Suspended solids 1 000 mg/l Total dissolved solids at 105ºC 4 000 mg/l Chloride as Cℓ 1 500 mg/l Total sulphates as SO4 1 500 mg/l Total phosphates as P 25 mg/l Total cyanides as CN 20 mg/l Total sulphides as S 50 mg/l Phenol index 50 mg/l Total sugars and starches as glucose 1 500 mg/l Oils, greases, waxes and fat 400 mg/l Sodium as Na 1 000 mg/l Case Study – Effluent Metering All of the parameters in Section A can be continuously monitored in order to determine not only compliance with the by-laws but also times where product has been introduced into the effluent system. A company in Cape Town was constantly in breach of the local discharge limits pertaining to TDS and pH and as a result installed continuous monitoring equipment to verify the council bills. The metering EFFLUENT not only indicated that there was significant variation in the discharge quality over MONITORING the month but that a number of isolated cases had a major impact on both the The metering conductivity and the pH. As product was dumped into the effluent system the TDS would increase and the pH would drop. The metering system allowed the company system allowed the to identify the day and time of the incidents and rectify the issues accordingly. company to identify the day and time of A composite sampler was installed so as to ensure that the municipality’s sample the incidents and was representative for the month rather than coinciding with one of the process rectify the issues lapses. A further indirect benefit was the ability to reduce raw material and product accordingly losses which were the cause of the effluent quality discharge spikes. 33 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 18 000 12,00 16 000 10,00 Total Dissolved Solids (TDS) 14 000 12 000 8,00 pH 10 000 6,00 8 000 6 000 4,00 4 000 2,00 2 000 - - TDS pH Figure 14. Effluent Quality Monitoring Project. A magnetic flow water meter was also installed and the actual discharge was monitored over a 12 month period. Council’s billing department calculation assumed that 95% of the water consumed was discharged to effluent. The actual metering indicated that this was only 79% (see table below). Table 9. Overview of water utilised and discharged over the course of a year. Total usage – kl 46 115 Total discharge - kl 36 430 % Discharge 79% The revised figure was submitted to council and their billing system adjusted accordingly. The company was also able to back-date the savings (>R1Million) from the beginning of the monitoring period which more than covered the costs for implementing the metering (10%), condensate monitoring and optimisation systems can be based on historical trends if a plant has direct steam injection processing systems in use. If no direct steam injection systems are in place and impurities in condensate are not present, then a good target for condensate return would be anything in excess of 85%. 6.3.4 Generation Efficiency A further important step would be to determine the generation efficiency using the mass balance approach which would typically be applied to monthly data collected for the boiler systems. This can be done by using steam tables in order to understand the energy content in the boiler feed and in the boiler steam which would then be used in conjunction with the fuel energy utilised for the system. The equation below summarises the calculation. Boiler Gneration Efficiency = Steam energy Content-Feedwater Energy Content Fuel Energy Content Flue gas analyses, while useful, should always be read in conjunction with the overall generation efficiency as isolated measurements do not necessarily capture losses related to low load conditions or start up. Case Study – Steam System Optimisation Klein River Cheese (Standford – Western Cape) utilized a 4 tonne / hr boiler to supply their processes. A steam assessment conducted by the National Cleaner Production Centre was conducted and the process steam load was calculated indirectly based on the boiler feed and blow down TDS readings. The generation efficiency of the steam could be calculated based on the calculated steam production and the boiler fuel utilised and was found to be low at 53%. Once the system losses were included, the overall system efficiency was in the vicinity of 33%. This should theoretically be over 75%. Condensate losses Direct Steam Area Boiler 52.9% Application Boiler losses Radiation loss / Boiler design* 2.0% Condensate return Direct steam Flash steam 12.8 tonne 0.0 tonne 0.8 tonne Flue gas loss 16.7% Heat Exchange Applications Unaccounted for loss 27.7% Bleed* 0.6% Generation efficiency Steam 53.0% 27.2 Tonne Generation efficiency 53.0% Hotwell Steam losses Radiation loss 21.3% Boiler feed 28.0 kl Leakage 5.6% Boiler Condensate return loss 8.5% Flash steam loss 2.4% Make-up Heat recovery 0.0% 16.0 kl Blowdown Sub total 37.8% 0.8 Tonne Cycles 35.7 Overall thermal efficiency 33.0% Bleed rate 2.8% Figure 17. Klein River Cheese steam system generation and system efficiency. 37 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA The solution was to revert to a point-of-use hot water system even though the input fuel costs were more expensive than the previous boiler fuel source. As a whole, significant savings were realised in addition to reduced maintenance costs as the steam system was decommissioned. The annual savings amounted to R290 000 (50% saving) with an investment cost of roughly R100 000. 6.4 Electrical Metering Some common terms used in electrical metering and important for understanding utility bills are provided below. Volt (V): The measure of electric potential between two points in a circuit and typically measured with a voltmeter or potential transformer. Apparent Power (Volt- Ampere - VA): The measure of “apparent” rate of energy supplied to an electric load. The volt- ampere (designated VA) is defined as the voltage multiplied by the current. The volt-ampere is the metric used to rate many forms of electrical equipment. Real Power (Watt): A measure of the “real power” delivered to an electric load. Watts are defined as volt-amperes multiplied by the “power factor”. As such, the real power will always be less than or equal to the apparent power. Real power is sometimes referred to as “active Power” or “true Power”. Reactive Power (Volt-ampere reactive - VAR): A measure of the system’s reactive power – or power stored in a system’s inductive or capacitive loads – and is mostly used for identifying power factor correction needs. Power factor: The ratio of “real power” (watts) to “apparent power” (volt-amperes) and defined as the cosine of the phase angle between voltage and current. For resistive loads (in ac circuits), the voltage and current are in phase and, therefore, the cosine of the angle is unity (i.e., 1.0), resulting in a power factor of unity. For loads with reactive components (e.g., motors, electrical ballasts), the voltage and current are not in phase resulting in a power factor of less than unity. Power factors significantly less than 1.0 (e.g., 0.85) can result in surcharges from the utility due to their need to make up the balance resulting from the improper power factor. The significance of power factor is that the electric utility supplies customers with Apparent Power (VA) but bills the customer for Real Power (Watts). A power factor below 1.0 requires the utility to generate more than the minimum volt-amps to supply the load. Demand: A measure of the average real power over a specified time interval. Depending on the utility, the specified interval is between 5 minutes to 1 hour, with the 30-minute interval being the most common. Maximum demand: The highest average demand measured in kVA or kW at the point of supply during a 30 minute integrating period in a billing month. It is important to understand how your utility assesses maximum demand, and the associated kW charge, to be able to manage for operational and economic efficiency. Harmonics: A measure of the electrical frequencies beyond the fundamental frequency of 50 hertz and usually labeled as the first harmonic (50 hertz), second harmonic (100 hertz), and so on. Harmonics are created by non-linear loads (e.g., computer power supplies, electronic ballasts) that draw current in short pulses rather than the traditional smooth ac sine waveform. Among other problems, harmonics can cause excessive heating of metal wires and certain types of electrical interference. Total harmonic distortion (THD): THD is a measure of the content of all major types of harmonic frequency current or voltages in relation to the fundamental current voltage frequency. This content is usually expressed as a percentage of the fundamental frequency and is defined as the square root of the sum of the squares of the harmonics divided by the fundamental frequency. 6.4.1 Meter Selection Typically monitoring of power is based on supplementing strategies to reduce energy costs (i.e., maximum demand strategies) as well as verifying monthly bills. Increasingly, with the adoption of energy management systems, digital electrical metering is being rolled out to departments and to large electrical energy users as the norm in plants. The installation cost relative to the ability to obtain a large amount of useful information as well as control equipment can be considered to be very low when compared to water and steam metering applications. The table below provides 38 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA an indication of the advantages and disadvantages of the different types of meters but with the cost of digital meters coming down significantly the only consideration would be for Mechanical or Electro-mechanical meters if they feed into an existing SCADA system. Table 11. Overview of electrical energy meters and their characteristics. Advantages Disadvantages Application Mechanical • Low Cost • Typically manually read • Monitor utilisation on air / • Reasonably Accurate • No time based recording cooling compressors • Not able to monitor quality Electro- • Low Cost • Not able to monitor quality • Remote monitoring Mechanical • Reasonably Accurate • Separate data logger applications (i.e. borehole) • Can produce a pulse output needed to collect • Large Energy Users for logging information (compressors) Digital Meters • Accurate • Moderate to high cost • Site Feed • Data storage with time • More complicated data Departments stamp management requirement Large power users / areas • Accommodate multiple • Additional systems for data inputs transfer and use • Two way communication • Built in alerts • Flexible data intervals Case Study – Compressed Air A blow moulding operation in Gauteng supplying HDPE and PET bottles to the dairy industry utilised a 200kW and 250kW compressor for their 700 psi line and alternated the compressors on a weekly basis for maintenance purposes. Power, flow and pressure monitoring equipment was installed over a period of two weeks. When the 250kW compressor was utilised there was a 30kW increase in power consumption with no significant change in the pressure or flow requirement. Utilising the more efficient compressor would realise a 130 000kWh saving per annum (~ R200 000 /annum) for the plant. Power Drawn Pressure pre-receiver Pressure post-receiver Compressed air flow Figure 18. Logging data for the compressed air system including power, pressure and flow. 39 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 40 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 7. METERING COMMUNICATIONS AND STORAGE Traditional metering systems in South Africa still dominate the Agri-processing industry with the bulk of non-electrical metering still being manually collected and then typed out onto a spreadsheet for review. For those companies who aggressively collect data, the opposite extreme exists where the information is in a format that is hard to analyse (e.g. 10 second data) or difficult to extract from the existing databases. The figure below summarises the different approaches to data collection and storage. METERING COLLECTION STORAGE Bills Manual On-site server Meter Data logger On-line server Variables Software Figure 19. Overview of data collection approach. Some key considerations should be kept in mind when considering the data collected: Manual Reading – Data integrity issues as will be discussed below. Automated Reading - Equipment should be fitted with automated reading options. These could be in the following outputs. a) Analog output – typically, 4 to 20 mA or 0 to 5 volts dc b) Contact closure – pulse type output c) Digital output – digital pulse d) Digital signal – outputs using networked communications (e.g., Ethernet, Modbus, HART). Storage and storage period – high sample rates stored for long periods will see high data storage requirements. Proprietary Communication Protocols – most companies have a range of different types of meters across the plant and care should be taken to avoid installing meters that make use of proprietary and encrypted communication protocols. This will make centralised collection of data at some point in the future costly as bespoke systems will need to be installed. 41 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 7.1 Common Mistakes Made Week/Year Date Total KWh KWh used 46 17/11/2013 410517 12484 Care should be taken to avoid common mistakes in collecting data. These include: 47 24/11/2013 422377 11860 • The date 02/02/20 could be the 2 February 2020 or the 48 01/12/2013 432119 9742 20 February 2002. Care should be taken to ensuring 49 08/12/2013 44 3846 11727 the date stamp is not ambiguous. For example, it is 50 15/12/2013 455635 11789 difficult to ascertain whether the data points to the right indicate the start of the week or the end of the 51 22/12/2013 467424 11789 week. 52 29/12/2013 Avoiding rows in the middle of a data set • 540433 (discontinuous data) that are not a part of the data set needing to be analysed. These will need to be 1 05/01/2014 475490 8066 removed before analysing. From the table on the left 2 12/01/2014 487088 11598 the weekly data for 2013 is summed as well as the 3 19/01/2014 498602 11514 monthly data for January and is included in the data 4 26/01/2014 510154 11552 set. • The data intervals / periods should be consistent. 517633 7479 This can frequently be a problem when working with 5 02/02/2014 520624 2991 municipal bills which have differing number of days 6 09/02/2014 533721 13097 in each billing month that can range for 25 to 31 depending on the billing cycle. In this data set, the 7 16/02/2014 541942 8221 point 28/02/2014 has only 5 days in the week (starting 8 23/02/2014 554224 12282 23/02/2014). 28/02/2014 564103 9879 • Estimates should be avoided and omitted from the 9 02/03/2014 568055 3952 data set before analysing. In this example, the data points for the 16 and 23 March 2014 were not taken 10 09/03/2014 578027 9972 and a zero figure was used in the data set which 11 16/03/2014 0 would significantly skew any analyses conducted . 12 20/03/2014 0 • Finally, if data is manually entered, the data readings 13 30/03/2014 589386 11359 should be taken at the same time of day. 7.2 Best Practice For Datasets Datasets should ideally be: • Automatically collected. • Have reporting options that allow output with set time intervals including: »» 30 minute »» 60 minute »» Daily »» Weekly »» Monthly »» Annually • Automated reporting and alerts. 42 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 7.3 Case Study – Power Factor An energy audit was conducted at a food plant based in Johannesburg. Power factor correction equipment was installed but the unit had tripped. The resulting power factor of 0.86 equated to an additional demand of roughly 120kVA which would amount to R20 000 / month additional cost. This plant was well run and had installed continuous logging systems but had not set up monitoring systems which would notify management when conditions were not in specification. In this instance the cost of the kVA excess to the plant was in the vicinity of R40 000 for the two months that the trip went undetected. This shows the importance of not only metering key points effectively but ensuring that a monitoring programme is implemented with feedback mechanisms in place to notify staff of conditions out of the operational specifications. Figure 20. Demand graph illustrating the difference between kVA supplied and kW drawn when the power factor equipment had tripped. 43 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 44 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 8. METERING COSTS AND FINANCING OPTIONS The costs for meter installation for most of the systems found in the Agri-Processing sector are provided in the sections below. These costings assume that the site is in a major metropolitan area and that the installation has no practical complications relating to the installation. The costings are also an indication of the pricing as of November 2021 and should be escalated when used. Ongoing maintenance and calibration can be budgeted annually at 10% of the cost of equipment installed. 8.1 Water Metering In many cases there are existing water meters in place that can be used to capture continuous flow data. Should this not be the case then the table below can be used as a budget price for replacement of the meters or for a new installation. Many companies manually read the meters and compile the information on spreadsheets. We would recommend automating this process and a budget for this is also provided in the table below. The monthly subscription would apply should the data be hosted by a third party service provider. Table 12. Budget price for the installation of a water meter.6 Nominal Diameter (MM) Installed Cost Data link per point Monthly subscription per point 15 R 3 400 R 3 000 R 60 20 R 3 600 R 3 000 R 60 25 R 4 000 R 3 000 R 60 40 R 4 400 R 3 000 R 60 50 R 7 600 R 3 000 R 60 80 R 10 000 R 3 000 R 60 100 R 15 000 R 3 000 R 60 150 R 25 000 R 3 000 R 60 200 R 50 000 R 3 000 R 60 250 R 75 000 R 3 000 R 60 300 R 125 000 R 3 000 R 60 400 R 250 000 R 3 000 R 60 450 R 400 000 R 3 000 R 60 6 Personal Communication (August 2021) G.W. TRAUTMANN BK and Livewire Engineering (www.livewire.co.za) 45 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA Table 13. Budget price for the installation of quality meter types. Metering Point Installed Cost Data link per point Monthly subscription per point Effluent - pH R 15 000 R 3 000 R 60 Effluent - TDS R 15 000 R 3 000 R 60 Effluent - COD R 30 000 R 3 000 R 60 Custom Composite Sampler* R 100 000 Total R 160 000 * Custom composite sampler using flow to activate a peristaltic pump discharging into a refrigerated container with a manual overflow pipe fitted. 8.2 Electrical Energy Metering Electrical meter costs would vary depending on the size of the installation and the number of feeds into the plant. As a budget cost for a 1MVA installation, the following costs would apply. Table 14. Budget price for installing an electrical energy meter.7 Item Cost Meter R 7 500 Data Link R 5 000 Installation R 2 500 Total R 15 000 Monthly subscription per metering point R60 8.3 Compressed Air Systems Compressed air systems are a lot more costly to monitor as they require data measurement intervals in the second range and the data logger requires multiple inputs with a visualisation screen for trouble shooting at the compressor room itself. Table 15. Budget Cost for installing a compressed air system metering programme.8 Item Cost Data logger with visualisation capability* R 75 000 Compressor monitoring (power, flow, pressure and moisture) R 75 000 Remote pressure and flow monitoring point linked to data logger R 30 000 Total R 180 000 Monthly subscription per metering point R60 * Not required if data is captured to a SCADA system 7 Personal Communication (August 2021) Livewire Engineering (www.livewire.co.za) 8 Personal Communication (August 2021) CS Instruments (www.cs-instruments.co.za) 46 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 8.4 Refrigeration / Chiller Systems Cooling compressors can be cost-effectively monitored using conventional electrical and temperature metering systems although the temperature probes should be of a good quality and regularly calibrated. Flow data can be determined from pump curves and design data should flow metering prove to be cost prohibitive. Table 16. Budget Cost for installing a chiller system metering programme. Item Cost Electrical meter per compressor x 2 R 20 000 Temperature probes for feed and return lines per cooling loop x 2 R 10 000 Pressure probes for refrigerant lines x 2 R 10 000 Flow meter per cooling loop R 20 000 Data link for compressor room R 10 000 Total R 70 000 Monthly subscription per metering point R60 * Optimisation software not included in the costings above. 8.5 Boiler Systems The boiler system costing assumes the indirect approach to metering will be utilised which would require the installation of an automated TDS blowdown controller. TDS controllers will usually realise operational and maintenance cost savings as the water treatment controls are improved. The TDS controller and O2 probes can be excluded from the monitoring system as the main objective will be to measure generation efficiency which can be done with the feed meter and fuel consumption figures, as well as spot checks on the boiler blowdown TDS. The table below assumes a two boiler configuration in the plant. Table 17. Budget Cost for installing a steam system metering programme. Item Cost TDS Controller x 2 R 100 000 Temperature Probe x 2 R 10 000 O2 or CO2 probe x 2 R 100 000 Hot Water feed meter – 80mm R 20 000 Make up meter – 40mm R 5 000 Liquid fuel flow metering system – 40mm R 20 000 Boiler House data link / logger R 5 000 Total R 260 000 Monthly subscription per metering point R60 * Optimisation software not included in the costings above. 47 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 8.6 Overview Budget The following budget can be utilised for a plant with the following requirements: • Refrigeration plant with two chillers • Two boilers • One municipal water feed and five sub-metering points requiring continuous logging • One electrical feed with five sub-metering points • One compressed air system with two compressors and one remote logging point • An effluent monitoring point (TDS, COD and pH) with a composite sampler Comprehensive metering programmes will usually have a payback of less than three years. A budget for this sort of programme is provided in the table below. Table 18. Budget Cost for installing a comprehensive metering programme. Item Cost Water Metering Programme (6 points) R 100 000 Electrical Metering Programme (6 points) R 100 000 Compressed air plant with one remote point (10 points) R 180 000 Refrigeration Plant (7 points) R 60 000 Boiler plant (9 points) R 260 000 Effluent monitoring point and composite sampler (4 points) R 160 000 Total Capital Costs R 860 000 Monthly subscription for hosted solution R2 500 48 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 49 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 9. DATA ANALYSES AND USAGE The previous sections reviewed the common metering types and locations in a plant but attention must be given to interpreting the data that is generated. The following sections review common performance measurement approaches and introduce statistical analyses as a best practice tool for reviewing data. 9.1 Performance Measurement Performance metrics are historically determined by simple metrics or ratios primarily as a result of finance departments needing to set budgets and assess performance against the set budgets. The different types of metrics are discussed in section 3.2. This section will look to expand on the concepts and provide tools to be able to analyse data more effectively. 1. Simple trend of monthly electricity use Simple assessments of trends are not easy to judge from month to month especially if the process is seasonal. They can be useful to use if production is stable (little variation in production) volumes. The figure to the right is a graph of the plants total energy consumption per month likely derived from billing data. One will note the seasonal variation with peaks during the winter period. 2 000 000 1 800 000 1 600 000 kWh per month 1 400 000 1 200 000 1 000 000 800 000 600 000 400 000 200 000 0 01-2009 04-2009 10-2009 01-2010 04-2010 07-2010 01-2011 04-2011 07-2011 10-2011 04-2012 07-2012 10-2012 01-2013 07-2009 10-2010 01-2012 04-2013 Figure 21. Monthly kWh data for a factory. 2. Annualised energy trend A method to negate the impact of seasonality is to utilise an annual rolling average. Increasing or decreasing trends can be easily identified over a number of years. These trends do not incorporate relevant variables that may influence production. The rolling average for each month is the sum of the previous 12 months (including the month in question) divided by 12. 19500000 kWh per year (ELEC) 19000000 18500000 18000000 17500000 17000000 12-2009 03-2010 06-2010 03-2011 06-2011 09-2011 12-2011 03-2012 06-2012 09-2012 12-2012 09-2010 12-2010 03-2013 Figure 22. Annual rolling average of the same data. 50 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 3. Annualised energy and cost trends A difficulty with savings initiatives is that, while there may be a saving in energy or water consumption, the actual financial savings may not be seen if the rates increase exceeds the actual energy savings realised. This can often give the sense that the savings programme is not effective. The savings reflected by the rolling average for kWh is not translated into the rolling average for costs for the same period. $ 19 500 000 $ 1 400 000 Consumption Cost $ 1 200 000 kWh per year (ELEC) $ 19 000 000 $ 1 000 000 $ 18 500 000 $ 800 000 $ 18 000 000 $ 600 000 $ 400 000 $ 17 500 000 $ 200 000 $ 17 000 000 $0 12/2009 03/2010 06/2010 09/2010 12/2010 03/2011 06/2011 09/2011 12/2011 03/2012 06/2012 09/2012 12/2012 03/2013 Figure 23. Annual rolling average and cost. 4. Specific ratios A common method to factor in the impact of increased or decreased production is to utilise a ratio between the resource consumed and the amount of production. In this example the total kWh / month divided by the total production (in kg) for the same month. While this method accounts for changes in production, it does not take into account the effect of seasonality or baseload (consumption not related to production) and as a result have distortions in low production months or seasons.  5,5  5,0  4,5 kWh / kg Product  4,0  3,5  3,0  2,5  2,0 Figure 24. Ratio of kwh to production. 51 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 5. Statistical Tools The material used in this section draws on the National Cleaner Production Centre South Africa’s (NCPC-SA) two day course on Performance Measurement Indicators which goes into more detail and rigour on using statistical models. A linear regression model can be developed which develops a predictive formula for resource usage performance given one relevant or influencing variable. Most spreadsheet programmes provide the option to insert a linear regression from a standard X-Y plot when developing a chart. MS Excel also provides formulas to provide the values without having to generate the chart as indicted in the slide to the right. The model provides an equation (y=Mx+C) which allows one to calculated consumption given a level of production (or other relevant variable). The equation below would apply. Expect Consumption = Production x M (slope model constant) + C (y-axis intercept model constant) Scatter Diagram 1 600   You can also use formulae in excel 1 400 ü  c: =INTERCEPT (known_y's,known_x's) 1 200 ü  m: =SLOPE (known_y's,known_x's) 1 000 ü  R2 =RSQ(known_y's,known_x's) kWh/week 800 600 ü  Remember: y= Mx+C 7  C and m are constants 400 7  X is a measured “relevant variable” 200 y = 18,572x + 167,84 R² = 0,8926 0 0 10 20 30 40 50 60 variable Figure 25. Example of a scatter plot with linear regression through the data points. The statistical model provides an indication of the expected resource consumption should the production (or relevant variable) be zero. This is indicated in the slide to the right as the point where the X axis is zero. Examples of contributors to baseload are: • Lighting systems • Radiation losses • Water leaks • Cold storage In any efficiency programme, focus should be on optimising the efficiency of the programme (i.e. the M in the y=Mx+C equation) as well as reducing the baseload (the C in the y=Mx+C equation). Scatter Diagram 1 600 1 400   Intercept: 1 200 1 000   Consumption when the variable is 0. kWh/week 800   It can be considered as the 600 baseload in most of the cases. 400 200 y = 18,572x + 167,84 R² = 0,8926 0 0 10 20 30 40 50 60 variable Figure 26. Illustrates how to determine the plants baseload. 52 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA Statistical tools provide an indication of how much confidence one can put into the formula generated. One of these is the Coefficient of Determination or R-Square correlation (R2) which is a value between 0 and 1. The R2 value indicates how much of the variation in the data can be explained by the influencing variable used. A high R2 value (>0.75) indicates that the equation can be used to predict consumption. Conversely, a low R2 (<0.5) also provides an indication poor control and internal inefficiencies or the possibility that the selected independent variable is not the correct or sole influencer of the the resource consumption. Scatter Diagram 1 600   R2: 1 400   % of variation explained by variables 1 200   High R2: 1 000 kWh/week a) Strong correlation. Not necessarily good 800 performance. 600   Low R2: 400 a) There are other variables. 200 y = 18,572x + 167,84 R² = 0,8926 b)  Saving Opportunities in operational control. 0 0 10 20 30 40 50 60 variable Figure 27. How to determine the R2 value. 9.2 Case Study Data over a period of time can be difficult to analyse especially if the production is seasonal and the energy or water consumption is strongly influenced by external weather conditions. The table below is of an actual plant’s coal usage and production in 2020. The plant utilises the simple ratio (tonne Coal / tonne production) to measure performance. Anything below 1 is deemed to be good performance whereas everything above 1 is deemed to be poor. Table 19. Overview of 12 month data for the plant.   Coal Consumption Production Heating Degree Days (HDD) Simple Ratio (Tonne / month) (Tonne / Month) (13.5 °C) Jan-20 969 1 032 0 0.9 Feb-20 1 021 1 235 0 0.8 Mar-20 868 949 0.4 0.9 Apr-20 0 0 11.5 Err May-20 798 328 64.3 2.4 Jun-20 943 752 129.4 1.3 Jul-20 1 441 1 249 115.9 1.2 Aug-20 1 184 1 207 70.1 1.0 Sep-20 1 085 1 259 13.1 0.9 Oct-20 1 221 1 265 1.9 1.0 Nov-20 1 049 1 253 0.7 0.8 Dec-20 644 741 0 0.9 Average 935 939 33.9 1.0 53 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA One will note the impact of the COVID lockdown on their production volumes over the period April 2020 – Jun 2020. One would typically expect the performance to be poor when using intensity targets. In this case the intensity target could not be used as the there was no production in April 2020. July 2020 stands out though as it is a high production month yet it still performs poorly based on the intensity target method. Similarly, December 2020 stands out as being a good performing month even though it is the second lowest production month in the year. Conventional intensity targets are useful for quick assessments but are significantly distorted when a plant has a high baseload (non-production consumption) or if there are factors that strongly influence energy consumption that are not directly related to production. Statistical tools are a better way of measuring actual performance. In this instance, the company generated a simple X-Y plot in excel and opted for a liner regression. 1 600 y = 0,7599x + 221,71 1 400 R² = 0,77103 1 200 Tonne / month 1 000 800 600 400 200 0 0 200 400 600 800 1 000 1 200 1 400 Production Figure 28. X-Y plot with a linear regression. The statistical tool has determined the baseload to be 221 tonne coal per month (fuel consumption not dependent on production) and the simple ratio of 0.76 tonne coal / tonne production can be an effective measure of performance once the baseload has been taken into account. In this case, the baseload can potentially be ascribed to radiation losses, firing losses and leaks (condensate and steam). The equation provided below calculates the expected coal consumption once the baseload has been taken into account. Expect Coal Usage (Tonne) = 0.77 x Production (Tonne) + 221 tonne coal (baseload) While the above regression provided a better method for measuring efficiency, other factors had a significant impact on coal consumption. From the table, it was clear that the winter months tend to be the poorer performing months. An online tool (www.degreedays.net) was used to determine the impact that average ambient temperatures have on coal consumption. The site expressed this in the form of Heating Degree Days (HDD) which is a measure of how much (in degrees Celsius), and for how long (in days), the outside air temperature was below a certain level. The data was used to compile a regression with more than one relevant variable (HDD and production) and an equation was generated. The new model had a statistically more reliable prediction of expected coal consumption based on weather and production. Expect Coal Usage (Tonne) = 0.78 x Production (Tonne) + 2.78 x HDD+106 tonne coal (baseload) In the following table, tThe cells highlighted in green depict better than expected performance when using the different models. One will note that the simple ratio and the statistical model only using production as a relevant variable misinterpret performance specifically in summer months. The regression model using production and average ambient temperatures provides a very different view on the better performing months. 54 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA Table 20. A plant’s 12 month data including regression analyses outputs. Production Production and Weather   Coal Production HDD Simple Expected Difference Expected Difference Usage [tonnes] (13.5 oC) Ratio Coal [tonnes] Coal Usage [tonnes] [tonnes] [Coal Usage [tonnes] Usage / [tonnes] Production] Jan-20 969 1 032 0 0.9 1 006 -37 913 55.6 Feb-20 1 021 1 235 0 0.8 1 160 -139 1 072 -50.9 Mar-20 868 949 0.4 0.9 942 -75 849 18.6 Apr-20 0 0 11.5 Err 222 -220 139 -137.1 May-20 798 328 64.3 2.4 471 328 542 256.7 Jun-20 943 752 129.4 1.3 793 150 1 055 -111.6 Jul-20 1 441 1 249 115.9 1.2 1 171 271 1 405 36.0 Aug-20 1 184 1 207 70.1 1.0 1 139 45 1 245 -61.2 Sep-20 1 085 1 259 13.1 0.9 1 179 -94 1 128 -43.0 Oct-20 1 221 1 265 1.9 1.0 1 183 37 1 101 119.5 Nov-20 1 049 1 253 0.7 0.8 1 174 -125 1 088 -39.3 Dec-20 644 741 0 0.9 785 -141 686 -42.0 Regardless of what tool is used, data should never be utilised without a practical understanding of the influencing factors. The utilised coal consumption figures are based on purchasing records rather than actual consumption which could lead to inaccuracies in the input data. To this extent, a model was developed utilising the boiler feed water supply (indirect measure of steam generated), production and Heating Degree Days. The model provided an acceptable correlation with a high degree of data accuracy and 52 data points (weekly) instead of 12 (monthly). While utilising regression data will be able to provide insight into the process controls as well as a predictive tool for setting realistic targets taking all influencing variables into account, additional detailed process and system analyses should still be conducted to understand if efficiency opportunities exist as well as the savings potential. The United States Department of Energy Open Source software tool called “MEASUR” has many useful functions for analysing both historical data and trends as well as system data (i.e. steam, pumps, fans and compressed air systems). https://www.energy.gov/ eere/amo/measur The Website www.degreedays.net provides a useful platform to obtain weather data for various locations across the world. The web-site also provides a regression tool which you can paste data and it will find the best temperature set- points for a regression. 55 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 56 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 10. CASE STUDIES 10.1 Atlantis Water Supply Scheme The Atlantis Water Supply Scheme (AWSS)9 and the associated groundwater recharge scheme is a good example of how improved monitoring has allowed for sustainable water extraction from the groundwater reserves. The scheme was initially SUSTAINABLE WATER augmented with treated waste water and storm water however the quality of the EXTRACTION two streams resulted in a steady increase in the salinity of the ground water. The Implementing waste water stream was subsequently excluded from the augmentation scheme but as a result there was a reduction in ground water reserves. Treated domestic effective metering effluent was included and then subsequently, low salinity industrial streams. The and monitoring interventions not only resulted in an increase in the water levels but also resulted in systems is crucial for a decrease in the overall salinity concentration. While the case study relates to a long term sustainable municipal scheme, the learnings are important for companies interested in long term management of sustainable extraction of resources from a water catchment area, and especially if resources. ground water is utilised in lieu of limited municipal potable water supplies. 200 All Wastewater and Stormwater recharged Only Stormwater 150 recharged Domestic Wastewater and Stormwater recharged Low salinity supply 100 augmentation 50 A4 0 1981 1986 1991 1996 2001 2006 52 Only Stormwater Domestic Wastewater and recharged Stormwater recharged 50 North-eastern part of WL elevation, m 48 wellfield All Waste- water and Stormwater Low salinity supply 46 recharged augmentation 44 WP51 G33108 South-western part of wellfield WP9 42 1984 1989 1994 1999 2004 Figure 29: Atlantis Recharge Quality and Level Sampling Record 9 Department of Water Affairs (2010), Strategy and Guideline Development for National Groundwater Planning Requirements. The Atlantis Water Resource Management Scheme: 30 years of Artificial Groundwater Recharge, P RSA 000/00/11609/10 – Activity 17 (AR5.1) 57 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 10.2 ABI Premier Place – Pheonix (McComb, 2016) Amalgamated Beverage Industries (ABI) is the leading soft drink business in the international SABMiller plc group of companies. The ABI Premier Place manufacturing plant is approximately 20 years old and functions mainly as a returnable glass bottling (RGB) beverage facility. The beverages are filled into 300ml, 500ml and 1.25l bottles. The company had been on an aggressive water savings drive which had seen its specific water consumption drop from 5.5 in 2011 to 2.3 in 2013. One of the savings interventions over this period was the installation WATER RE-USE of a tank to recover the back wash and reject water in the water treatment systems. This water was used on non-product surface SYSTEM SAVINGS cleaning and applications (i.e., crate washing, lubrication systems and Interventions included CIP rinse manual cleaning). The initial investigation indicated that the recovered water recovery and improved water would be sufficient for the recovered water demeand. A meter recovered water utilisation. was installed on the municipal make-up line to the plant and it was found that the recovered water demand outstripped the supply, and IMPACT OVER 3 MONTHS in addition, that the municipal water supply control valve was faulty resulting in municipal water displacing the recovered water. Actual Cost Savings R190,697 Actual Water Savings 7,008 kl The faulty valve was repaired and additional water from the cleaning- Cost of Project R32,000 in-place (CIP) systems was recovered. After the intervention, no further municipal make-up water was required for the recovery Payback Period 1 Month tanks. Furthermore, a programme was implemented to improve the efficiency of recoverd water usage to reduce the demand to ensure no additional municipal water was required. Figure 30. Picture of the recovery and blending tanks. 58 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 10.3 RFG Foods – Groot Drakenstein RFG Food has 14 state-of-the-art production facilities and two farms – dairy and pineapple. The production facilities are equipped with modern technology and certified according to international standards. The RFG Foods complex in Groot Drakenstein is located in Pniel Road near Franshhoek outside Cape Town. The complex comprises of Ready Meals, Puree Plant, Dairy, RFG Head Office and the Ayreshire Stud Farm as well as Services Departments. An energy assessment previously carried out in January 2013 The methodology included compiling detailed electrical energy balance and identifying opportunities for increased optimisation. Areas for resource saving identified included a 25% reduction in electrical energy costs and an 18% reduction in steam generation costs. Savings in excess of R 4,000,000 per annum were identified. During the audit a detailed analyses was conducted on the electrical energy consumption over the period of a year. The existing RFG tariff from Groot Drakenstein Municipality was a standard consumption (kWh) and maximum demand (kVA) tariff. The option to switch tariffs was reviewed and a Time of Use (TOU) tariff in was adopted in July 2013. There were no capex requirements. The savings provided immediate payback and significant reduction in electricity costs. Being a 24 hour operation, RFG were able to derive significant benefit from the lower off-peak tariff as well as the lower kVA charges. RFG Electricity Cost Comparison R 900,000 R 800,000 R 700,000 R 600,000 R 500,000 R 400,000 Prev. Tariff Cost R 300,000 Current TOU Cost R 200,000 R 100,000 R- Jul-13 Aug-13 Sept-13 Oct-13 Nov-13 Dec-13 Jan-14 Feb-14 Mar-14 Apr-14 May-14 Jun-14 Figure 33. RFG Electricity cost comparison July 2013 - June 2014 59 METERING AND MONITORING GUIDELINE FOR AGRI PROCESSING SECTOR IN SOUTH AFRICA 11. REFERENCES 1. Eskom (August 2021), “Tariff History”, retrieved 20 August 2021, (https://www.eskom.co.za/CustomerCare/ TariffsAndCharges/Pages/Tariff_History.aspx) 2. M Capes and S Moolman (18 Jan 2021), “The Price of Water and Electricity in South Africa: A Tale of Two Tragedies”. Retrieved 5 October 2021 (https://www.poweroptimal.com/the-price-of-water-and-electricity-in-south-africa-a- tale-of-two-tragedies/#chapter) 3. United Nations Industrial Development Organisation (Oct 2013), “Practical Guide for Implementing an Energy Management System”, Retrieved 5 October 2021 (https://www.unido.org/sites/default/files/2017-11/IEE_EnMS_ Practical_Guide.pdf) 4. JE van Zyl (October 2011). “Introduction to Integrated Water Management”, Water Research Commission TT 490/11, Retrieved 5 October 2021 (http://www.wrc.org.za/wp-content/uploads/mdocs/TT%20490-11.pdf) 5. The Carbon Trust (May 2007) “Advanced metering for SMEs - Carbon and cost savings”, Retrieved 5 October 2021 (https://www.carbontrust.com/resources/advanced-metering-for-smes) 6. Gavin Graham (October 2015) “RFG Foods EnMS Case Study”, Retrieved 5 October 2021 (https://www. industrialefficiency.co.za/wp-content/uploads/2021/05/Rhodes_Food_Group_EnMS_Full_case_study_2015.pdf) 7. Darrin McComb (December 2016) “Amalgamated Beverage Industries (ABI)- Premier Place – Water Optimisation Case Study” on behalf of the National Cleaner Production Center (NCPC-SA) 8. Darrin McComb (December 2015) Klein River Cheese Steam System Optimisation Case Study, Retrieved 5 October 2021 (https://www.industrialefficiency.co.za/wp-content/uploads/2021/07/KleinRiverCheese_SSO_FullCS.pdf) 9. JP Genet and C Schubert (2013) “Designing a metering system for small and medium-sized buildings”, Schneider Electric White Paper, Retrieved 5 October 2021 (https://sfigroup.co.za/wp-content/uploads/2021/04/Schneider- white-paper.pdf) 10. Department of Water Affairs (2010) Strategy and Guideline Development for National Groundwater Planning Requirements, “The Atlantis Water Resource Management Scheme: 30 years of Artificial Groundwater Recharge”, PRSA 000/00/11609/10-Activity 17 (AR5.1) 11. GP Sullivan et al. (March 2015) “Metering Best Practices - A Guide to Achieving Utility Resource Efficiency”, Release 3, Pacific Northwest National Laboratory for the Federal Energy Management Program U.S. Department of Energy, Retrieved 5 October 2021 (https://www.wbdg.org/FFC/DOE/NLCRIT/mbpg_2015.pdf) 12. I van der Stoep et al. (December 2012) “Awareness creation, implementation plans and guidelines for management of sustainable on-farm and on-scheme water measurement”, WRC Project No. K5/1778//4, Retrieved 5 October 2021, (http://watermeter.org.za/wp-content/uploads/2019/08/TT-550-12.pdf) 13. I van der Stoep et al. (December 2019) “Guidelines for selecting measuring devices for irrigation water measurement: Pipe flow”, WRC Project No. K5/1778//4, Retrieved 5 October 2021, (http://watermeter.org.za/wp-content/ uploads/2019/09/Pipe-flow-guidelines.pdf) 60 IN PARTNERSHIP WITH Schweizerische Eidgenossenschaft Confédération suisse Confederazione Svizzera Confederaziun svizra Swiss Confederation Federal Depar tment of Economic Affairs, Ed c at i o n a n d R es ea rc h EAER State Secretariat for Economic Affairs SECO