With more than 1,500 citations in ranked journals and hundreds of thousands of downloads, EPANET is the benchmark software for analysis and design of pressurized water distribution networks. User-friendly, powerful and reliable, its public domain use has spread throughout the world. However, from an energy standpoint its capacity is limited to the point at which, in certain circumstances it can supply erroneous results. This is understandable because on the one hand its main aim was to model water quality, and on the other hand, because it was conceived and developed towards the end of the last century before people started talking about the water–energy nexus. Increases in the cost of energy and the need to limit greenhouse gas emissions, however, have made energy efficiency a primary and inescapable objective. As the transport of pressurized water is a big consumer of energy, it seems convenient for EPANET users, particularly for those who applied this software to pressurized irrigation networks, to understand, as far as energy is concerned, its weaknesses and limitations which, in the end, is the aim of this paper.

NOMENCLATURE

  • Piezometric head at node i at time tk (m.w.c.)

  • Piezometric head at the reservoir i at time tk (m.w.c.)

  • Piezometric head of the pump i at time tk (m.w.c.)

  • Friction energy in pipes for the simulation period (kWh)

  • Energy trough leaks for the simulation period (kWh)

  • Energy supplied by the reservoirs for the simulation period (kWh)

  • Energy supplied for the simulation period (kWh)

  • Energy wasted in the pumping station for the simulation period (kWh)

  • Number of demand nodes of the network (dimensionless)

  • Number of pipes of the network (dimensionless)

  • Number of reservoirs (dimensionless)

  • Number of pumps (dimensionless)

  • Rotation speed of the pumping unit using one variable frequency drive (r.p.m.)

  • Nominal rotation speed of the pumping unit (r.p.m.)

  • Leakage flow rate at node i at time tk (l/s)

  • Flow rate at line j at time tk (l/s) that finally is lost through leaks

  • Flow rate supplied by reservoirs i at time tk (l/s)

  • Flow rate supplied by pumping station i at time tk (l/s)

  • Consumed flow rate at node i at time tk (l/s)

  • Flow rate necessary to satisfy the users demand that circulates at line j at time tk (l/s)

  • Time in the steady state simulation (h)

  • Total time of simulation (h)

  • Emitter exponent (dimensionless)

  • Specific weight of water (N/m3)

  • Time interval of integration (Δt = tk+1tk)

  • Friction losses in line j at time tk (m.w.c.)

  • Efficiency in pumping station i at time tk (dimensionless)

INTRODUCTION

The first version of EPANET, version 1.1, was released in 1993 (Rossman 1993) and 7 years later, version 2.0 (Rossman 2000) appeared, which is still the latest version of this software package. Eleven years later, the imminent appearance of version 3.0 was announced (Rossman 2011), which for whatever reason has not been released. In fact, there is a recent and very welcomed international attempt led by Dr Dominic Boccelli (University of Cincinnati) who, saddled with the responsibility to update EPANET, has addressed an open letter to the international Water Distribution Systems Analysis (WDSA) community entitled ‘Announcement of an open source EPANET initiative’. If successful, it should be great news for all EPANET users.

The following analysis of the weaknesses and limitations of EPANET as regards energy, should therefore be referred to the present version, 2.0. It must be said that, from an energy standpoint, few changes were added to the original for the second version of the software. The most important aspect being the inclusion of schedule patterns for the cost of energy, and allowing the study of total energy expenditure, based on pumping schedule arrangements. Neither does the awaited version 3.0 promise improvements as regards energy; despite this, some enhancements were announced in the modelling of the evolution of water quality, in hydraulics and in the interface (Rossman 2011). There is just one energy collateral enhancement, which is the possibility of pump behaviour modelling with third degree polynomials, an improvement that should be welcomed because minor errors (see Table 6, pump efficiencies, periods 2 and 3.2, with pump rotating at its nominal speed) could be avoided. This will overcome another well known EPANET drawback: positive slopes are not allowed along the head–flow characteristic pump curve, but they do happen often in practice (particularly in axial flow pumps).

In order to understand the lack of attention EPANET pays to energy-based analysis, it is important to place the program in its historical context. In fact, the first article to expressly cite the water–energy nexus (Gleick 1994) is contemporary with version 1.1 of EPANET, while the first report quantifying energy linked to the urban water network (CEC 2005) is later than the actual version of EPANET, a report which makes the water–energy nexus a primary objective. The activity of EPANET's promoter, the Environmental Protection Agency (EPA), reflects this, because they develop programs and tools to this end. Those related directly to the water industry stand on there own (EPA 2008, 2012), while the programs with a wider scope (e.g. the promotion of efficient energy devices with corresponding labeling) are developed in collaboration with other institutions (EPA 2009). One of its latest publications (EPA 2013) demonstrates the attention that EPA currently pays to the water–energy nexus. This is understandable. In the USA, electricity use accounts for 25–40% of the operating budgets for wastewater (NYSERDA 2008), while drinking water and wastewater systems account for 3–4% of the total consumption (Eisenberg 2012). These figures are even higher in California (Wolff 2010) where just the transport of pressurized water accounts for 6% of total energy consumption (Water in the West 2013). As a matter of fact, in Europe the agriculture energy demand is on average 2% of the total energy consumption (MOE 2011), a figure that in the more agricultural Mediterranean, southern countries, such as Spain, rises up to 3% (Corominas 2010).

This paper describes and shows, through a case study, the energy weaknesses of EPANET. Lastly, in order to adapt the most globally used water network software to present needs, some improvements are suggested.

THE EPANET ROLE IN PRESSURIZED URBAN AND IRRIGATION SYSTEMS ANALYSIS

EPANET was conceived in a bid to track the evolution of water quality from the source (injection point) to the customer's tap (user point). This general interest allowed EPA to allocate public funds to develop a software package that, at the end, will compete with commercial software programs promoted by companies that supported the state with their taxes.

Hydraulic analysis, including the extended period simulation, is a necessary prerequisite in performing water quality analysis. In all, the EPA developed a complete software package to control water quality throughout the system and, at the same time, it models the hydraulic system behavior. A collateral result from the very first moment, attracted much more attention from the hundreds of thousands of EPANET's users, for example, the water network sectorization study is one of the most widespread uses of the program (Renaud et al. 2014; Di Nardo et al. 2014), while for quality purposes, the program is rarely used. In the irrigation field, where all agricultural constituents (such as nitrogen) are non-reacting, their movement and fate within the distribution network does not change at all.

Even though it was not conceived to help them, agricultural designers of pressurized irrigation systems immediately became aware that EPANET should be a useful tool to facilitate their daily work. User-friendly, reliable, powerful and zero cost, are strong reasons to attract professionals of a field (with much less economic potential, than those of the water industry) who can easily pay for powerful and updated commercial packages. Even not being conceived to simulate irrigation network behaviours on demand schedules (to perform such analysis other software packages, are more appropriate) (Calejo et al. 2008) it is the most used program in the world of agricultural engineering, specially in developing countries (Arora & Jaiswal 2013; Abdulrazzaq & Jahad 2013) owing to its accessibility, reliability and is completely free of cost.

However, as said before, from an energy standpoint, this referent software package has limited capability to the point where in certain circumstances it can even supply erroneous results. Energy issues being of increasing importance in the urban and agricultural areas, users must be fully aware of the EPANET shortcomings in this regard. In what follows, firstly, major and minor EPANET energy's drawbacks are listed, later some improvements are suggested and finally a case study of an irrigation network conceived to show the previously described handicaps, is presented.

MAJOR ISSUES

The errors which can mislead program users are firstly mentioned.

EPANET calculates incorrectly the efficiency of variable speed pumps and, thus, power

This error has already been the focus of a technical note (Marchi & Simpson 2013). Although, this issue is well documented, details of where the error lies and how it is transmitted to other results (power and energy expenditure) will be given in this paper to help engineers avoid it.

EPANET ignores the natural energy

The transportation and distribution of pressurized water are carried out at the cost of the elastic (water and pipe) energy. This energy can proceed from two sources, natural (or gravitational) supplied by tanks and shaft energy supplied by pumps. EPANET only recognizes the latter energy. There is no energy report on gravitational systems and it only accounts for the energy which the pumps supply in mixed systems (natural and pumped energy). Therefore, when the energy is supplied (partially or totally) by a tank, EPANET's results do not allow energy audits to be performed.

EPANET miscalculates pumping expenditure when the simulation is not carried out within 24 h

In practice, it is relatively common to find pressurized systems operating daily for less than 24 h. This applies to intermittent water supplies, a frequent scenario in developing countries and habitual in pressurized irrigation networks. If the number of daily operational hours, nt, is less than 24, EPANET increases energy expenditure to the factor of 24/nt. The error is repeated for all simulation periods which are not multiples of 24 h. In irrigation systems, water is typically supplied for less than 24 h per day. This error once more is evidence that shows the program was conceived for urban water networks where, except for the intermittent supplies of the developing countries, the network operates 24 h per day and 365 days per year and therefore, this error has no impact.

MINOR ISSUES

Energy aspects that should be considered in order to adapt EPANET to present day needs are mentioned below.

Concerning energy consumption

The kilowatt hours which are really consumed in pumping stations (the real power demanded and the energy effectively consumed) are not solely dependent on pumping efficiency. The efficiency of electric motors and of variable speed drivers, which in certain circumstances can cause significant energy loss should be considered; this is an important and well known issue (Burt et al. 2008).

Concerning tariffs

EPANET only contemplates a single value for power contract costs while actually, time of day rates and seasonal variations apply, as occurs with the price of energy.

Concerning energy intensity

EPANET calculates the energy intensity (kWh/m3) as referred to the volume supplied to the system. As it is more representative if referred to the registered volume (consumed by the users), especially in highly leaky systems, it is convenient to provide simultaneously both energy intensities.

Concerning energy expenditure in systems with several pumping stations

EPANET, without any limitation in the number of nodes and lines in the system, has formidable powers of calculation. Networks of a significant size can be fed by different pumping stations which are subject to different tariff regimes, given that they currently depend on the maximum contracted power. The programme, in its present state, contemplates a single tariff for the whole system.

POTENTIAL IMPROVEMENTS

The first required improvement should be the correction of the preceding drawbacks, which included the adequate head-flow pump modelling, irrespective of its characteristics. Besides this, some new capabilities could be incorporated. The description of two of the most relevant ones are as follows.

Energy assessment. Using the EPANET input data and their results, very valuable information can be obtained from an energy point of view, for example, the ideal and real network performance. By comparing these values, the system's margin of energy efficiency improvement can be estimated. Furthermore, assessing the relevance of topographical energy permits us to identify the pipes in which this energy can be recovered by installing Pumps as Turbines (PATs) or, if that is not an option, dissipating it by pressure reducing valves (Cabrera et al. 2015).

Energy audit. This can also be performed to discover where the system is more inefficient. The audit, which matches the energy entering and leaving the system, can identify and quantify the losses. These can be caused by friction (in pipes and valves), pumping station inefficiencies and leaks which are a sum of the energy embedded in the water losses and the additional friction in the pipes, a consequence of the greater flow required (Cabrera et al. 2010).

Currently, the Programmer's Toolkit API (Rossman 1999) allows these calculations to be performed but, not directly. As energy efficiency becomes more and more important, it would be desirable to automate all of these calculations.

CASE STUDY

Figure 1 and Table 1 correspond to a system designed to show the weak points of EPANET as mentioned above. It is an irrigation network divided into three sectors, each of them being active for 6 h. The third sector irrigates in two 3 h periods. The system operates for a total of 18 h a day. Both pumps are fed by a tank at a constant level (30 m).
Table 1

Pipe and node data

Pipe Length (km) Diameter (mm) Node Elevation (m) Base demand (l/s) Emitter coefficient (m3–α/s) Period of demand (h) 
100 0.5 250 11 10 10 0.121221 0–6 
111 0.5 200 12 20 0.026605 0–6 
200 1.0 300 21 45 10 0.212212 6–12 
211 0.5 175 22 50 15 0.058170 6–12 
212 0.5 250 23 50 10 0.108260 6–12 
213 0.5 175 24 55 15 0.083563 6–12 
300 3.0 300 31 75 20 0.390568 12–15 
311 0.6 200 32 80 20 0.056493 12–15 
312 0.5 300 33 80 20 0.098951 15–18 
313 0.6 200 34 85 20 0.053649 15–18 
314 0.5 250 35 85 20 0.098423 15–18 
315 0.6 200 36 90 20 0.091827 15–18 
   RES 30 – – – 
Pipe Length (km) Diameter (mm) Node Elevation (m) Base demand (l/s) Emitter coefficient (m3–α/s) Period of demand (h) 
100 0.5 250 11 10 10 0.121221 0–6 
111 0.5 200 12 20 0.026605 0–6 
200 1.0 300 21 45 10 0.212212 6–12 
211 0.5 175 22 50 15 0.058170 6–12 
212 0.5 250 23 50 10 0.108260 6–12 
213 0.5 175 24 55 15 0.083563 6–12 
300 3.0 300 31 75 20 0.390568 12–15 
311 0.6 200 32 80 20 0.056493 12–15 
312 0.5 300 33 80 20 0.098951 15–18 
313 0.6 200 34 85 20 0.053649 15–18 
314 0.5 250 35 85 20 0.098423 15–18 
315 0.6 200 36 90 20 0.091827 15–18 
   RES 30 – – – 
Figure 1

Network layout.

Figure 1

Network layout.

Level zero is assigned to the node with the lowest elevation (node 12). Service pressure is 10 m, the elevation of the tank 30 m, and the roughness of the pipes 0.1 mm. Table 1 depicts the network characteristics and Table 2 shows those of the pumps which guarantee minimum pressure at high elevations. Leaks are modelled through a node's emitters.

Table 2

Pump characteristics

PUMP Head (H)–flow (Q) curve Efficiency curve Operating schedule Operating characteristics Rotation speed (N/Nο
P1 H1 = 53.33–0.003704 Q2 η1 = −0.0164 Q2 + 1.8708Q + 19.929 6–12 h 61.13 l/s; 39.49 m.w.c. 
P2 H2 = 136–0.00369 Q2 η2 = −0.0096 Q2 + 1.6707Q + 11.126 12–15 h 52.09 l/s; 66.49 m.w.c. 0.75 
15–18 h 106.13 l/s; 94.43 m.w.c. 
PUMP Head (H)–flow (Q) curve Efficiency curve Operating schedule Operating characteristics Rotation speed (N/Nο
P1 H1 = 53.33–0.003704 Q2 η1 = −0.0164 Q2 + 1.8708Q + 19.929 6–12 h 61.13 l/s; 39.49 m.w.c. 
P2 H2 = 136–0.00369 Q2 η2 = −0.0096 Q2 + 1.6707Q + 11.126 12–15 h 52.09 l/s; 66.49 m.w.c. 0.75 
15–18 h 106.13 l/s; 94.43 m.w.c. 

Each pumping station is on for 6 h (the irrigation period). The highest elevated sector (sector 3) is fed by pumping station P2 equipped with a variable speed driver to adjust the system's pressure requirements. For the first 3 h, it rotates at 75% of its nominal speed, subsequently rotating at 100% (Table 2). The pumps are timed to switch on and off.

The daily water balance in each sector (with the apparent losses included in the amount of water consumed) is provided in Table 3. Leakages are held to be pressure-dependent, and modelled by nodal emitters (exponent α = 1.1) with coefficients which are proportional to the length of the pipes converging on it (see Table 1).

Table 3

Water audit

  Sector 1 (0–6 h) Sector 2 (6–12 h) Sector 3 (12–15 h) Total (0–18 h) 
Vinjected (m3/day) 737.42 1,320.41 1,708.77 3,766.61 
Vconsumed (m3/day) 648 1,080 1,296 3,024 
Vleakage (m3/day) 89.42 240.41 412.77 742.61 
ηvolumetric 87.9% 81.8% 75.8% 80.2% 
  Sector 1 (0–6 h) Sector 2 (6–12 h) Sector 3 (12–15 h) Total (0–18 h) 
Vinjected (m3/day) 737.42 1,320.41 1,708.77 3,766.61 
Vconsumed (m3/day) 648 1,080 1,296 3,024 
Vleakage (m3/day) 89.42 240.41 412.77 742.61 
ηvolumetric 87.9% 81.8% 75.8% 80.2% 

By simulating the day-to-day operation of the system (18 h), with a hydraulic interval of 1 h, the global balance between the power supplied and consumed (energy audit) is calculated by means of the results provided by EPANET and following the methodology previously described (Pardo et al. 2013). The calculation of the energies involved in the system is based on the equations shown in Table 4.

Table 4

Energies involved in the energy audit

Energy consumed kWh Useful energy Useful energy delivered to users  
Energy losses Leakage energy losses  
Friction energy losses  
Wasted energy in pumping stations  
Energy supplied kWh   Shaft energy (supplied by pumps)  
Natural energy (supplied by external sources)  
Energy consumed kWh Useful energy Useful energy delivered to users  
Energy losses Leakage energy losses  
Friction energy losses  
Wasted energy in pumping stations  
Energy supplied kWh   Shaft energy (supplied by pumps)  
Natural energy (supplied by external sources)  

From equations shown in Table 4, tp being the period for calculating the expressions, Equation (1) concerning the final balance results in: 
formula
1
The previous Equation (1) states that the energy supplied by reservoirs and pumps must be equal to energy delivered to the users plus the losses, because this system does not include compensation tanks (Cabrera et al. 2010). The results are shown in Table 5.
Table 5

Power balance

Period Power consumed (kW)
 
Power supplied (kW)
 
Useful power Power lost (leaks) Power lost (friction) Power lost (pumping station) Total consumed Natural power Average pump power Total supplied 
1 (0–6 h) 8.34 1.17 0.54 10.05 10.05 10.05 
2 (6–12 h) 32.44 7.28 1.95 8.76 50.43 17.99 32.44 50.43 
3 (12–18 h) 57.67 18.78 12.97 16.08 105.51 23.28 82.23 105.51 
3.1 (12–15 h) 35.66 10.85 2.79 8.05 57.35 15.33 42.02 57.35 
3.2 (15–18 h) 79.69 26.71 23.15 24.11 153.67 31.23 122.44 153.67 
Period Power consumed (kW)
 
Power supplied (kW)
 
Useful power Power lost (leaks) Power lost (friction) Power lost (pumping station) Total consumed Natural power Average pump power Total supplied 
1 (0–6 h) 8.34 1.17 0.54 10.05 10.05 10.05 
2 (6–12 h) 32.44 7.28 1.95 8.76 50.43 17.99 32.44 50.43 
3 (12–18 h) 57.67 18.78 12.97 16.08 105.51 23.28 82.23 105.51 
3.1 (12–15 h) 35.66 10.85 2.79 8.05 57.35 15.33 42.02 57.35 
3.2 (15–18 h) 79.69 26.71 23.15 24.11 153.67 31.23 122.44 153.67 

In every scenario, the power supplied (the sum of the natural and shaft power) is equal to that consumed, and the sum of the useful power plus losses (friction in the pipes, dissipated in leakages and lost at the pumping station).

Table 6 compares the results provided by the EPANET energy analysis with the results of Table 5. For the sake of simplicity, the cost of the energy is taken to be time-constant, with an average value of 0.1 €/kWh. The differences correspond to the underlined errors:
  1. In period 1 (the lowest elevation area is irrigated), all the supplied energy is natural. EPANET does not consider this power. The natural energy is omitted in all three scenarios. Additionally, when the pumps are not working, energy consumption is always nil.

  2. In period 2, in which sector 2 is irrigated, the error in the estimation of cost of the energy must be added to the omission of natural energy. It can be seen how EPANET over-estimates the expenditure to the expected factor (24/18).

  3. Lastly, all three errors are present in scenario 3. Natural energy is not considered and, during period 3.1 when the pump is rotating at 75% of its nominal speed, the pump's efficiency is not correctly calculated, translating an error to power and costs. However the ratio is no longer 4/3. It is higher (1.37) because, as EPANET's efficiency is inferior to the real value, it overestimates the power of the pump.

Table 6

Comparison of the energy analysis and EPANET results

Period Calculated analysis
 
EPANET analysis
 
Average power (kW) Natural power (kW) Shaft power (kW) Pump efficiency (%) Cost/day (€) Average power (kW) Average efficiency (%) Cost/day (€) 
10.05 10.05 – – – – – 
50.73 17.99 32.44 73.00 19.44 32.72 72.33 26.17 
105.51 23.28 82.23 80.58 49.34 84.78 76.21 67.82 
3.1 57.35 15.33 42.02 80.85  47.02 72.26  
3.2 153.67 31.23 122.44 80.30  122.57 80.18  
Period Calculated analysis
 
EPANET analysis
 
Average power (kW) Natural power (kW) Shaft power (kW) Pump efficiency (%) Cost/day (€) Average power (kW) Average efficiency (%) Cost/day (€) 
10.05 10.05 – – – – – 
50.73 17.99 32.44 73.00 19.44 32.72 72.33 26.17 
105.51 23.28 82.23 80.58 49.34 84.78 76.21 67.82 
3.1 57.35 15.33 42.02 80.85  47.02 72.26  
3.2 153.67 31.23 122.44 80.30  122.57 80.18  
According to EPANET, the average efficiency of pump P2 is 76.21, while the average efficiency for the operating conditions of the example is 80.58. EPANET does not correctly calculate the efficiency when the rotational pump's speed is different from the nominal one, simply because it does not consider the affinity laws. More detail can be seen in Marchi & Simpson (2013). The inaccuracy is shown in Figure 2. For the working point P3.1, EPANET's efficiency is 72.26%, the value obtained from a direct lecture on the efficiency curve. But, taking into account the laws of similarity that apply to a pump working at different rotational speeds, the correct result is 80.85%. And the same applies to the second working point P3.2.
Figure 2

Pump operating point at two different speeds.

Figure 2

Pump operating point at two different speeds.

Lastly, it should be highlighted that, for the water tariffs contracted at the several pumping stations, it is impossible for EPANET to apply differing energy (and power term) prices.

CONCLUSION

From the point of view of energy, the calculation module of EPANET (the worldwide reference programme in both urban and agricultural fields) presents notable deficiencies and shortcomings. They have been classified into two main groups, major and minor issues (see the corresponding sections) and practically demonstrated in the case study. That happens at a time when energy efficiency and its consequences (greenhouse gas emissions) are of utmost priority, EPANET's energy component should be updated and completed. So far, its contribution to the energy efficiency improvements of these systems can be substantially enhanced if the issues previously discussed are taken into account. EPANET (the program, the source and the toolkit) has played an invaluable role in practice and research and, with some minor rearrangements, improvement in its brilliant service record will follow.

REFERENCES

REFERENCES
Abdulrazzaq
K. A.
Jahad
U. A.
2013
Evaluation of uniformity coefficients for sprinkler irrigation system under traditional and looped network field conditions
.
International Journal of Computer Applications
68
(
21
),
40
44
.
Arora
S.
Jaiswal
A.
Kr.
2013
Optimal cost of irrigation network design using Epanet
.
International Journal of Computer Applications (0975–8887)
,
68
(
21
),
41
45
.
Burt
C. M.
Piao
X.
Gaudi
F.
Busch
B.
Taufik
N. F. N.
2008
Electric motor efficiency under variable frequencies and loads
.
Journal of Irrigation and Drainage Engineering. ASCE
134
(
2
),
129
136
.
Cabrera
E.
Pardo
M. A.
Cobacho
R.
Cabrera
E.
Jr
2010
Energy audit of water networks
.
Journal of Water Resources Planning and Management
2010
,
669
677
.
Cabrera
E.
Gómez
E.
Cabrera
E.
Jr
Soriano
J.
Espert
V.
2015
Energy assessment of pressurized water systems
.
Journal of Water Resources Planning and Management
.
doi: 10.1061/(ASCE)WR.1943-5452.0000494. 12 pp. Published on line on November 6, 2014
.
Calejo
M. J.
Lamaddalena
N.
Teixeira
J. L.
Pereira
L. S.
2008
Performance analysis of pressurized irrigation systems operating on-demand using flow-driven simulation models
.
Journal of Agricultural Water Management
95
(
2008
),
154
162
.
CEC (California Energy Commission)
2005
California's water- energy relationship
.
Final staff report
.
CEC 700–2005–011 SF California Energy Commission
.
State of California
.
Corominas
J.
2010
Agua y Energía en el riego en la época de la Sostenibilidad (Water and Energy in irrigation at the time of Sustainability)
.
Ingeniería del Agua
17
(
3
),
219
233
.
Di Nardo
A.
Di Natale
M.
Santonastaso
G. F.
2014
A comparison between different techniques for water network sectorization
.
Water Science & Technology: Water Supply
14
(
6
),
961
970
.
Gleick
P.
1994
Water and Energy
.
Annual Review of Energy and the Environment
19
,
267
299
.
EPA (Environmental Protection Agency)
2008
Ensuring a Sustainable Future: An Energy Management Guidebook for Wastewater and Water Utilities
.
January 2008
.
EPA, Office of Wastewater Management
,
Washington, DC
.
EPA (Environmental Protection Agency)
2009
ENERGY STAR Program Integrity Can Be Enhanced
.
Washington, DC
.
EPA (Environmental Protection Agency)
2012
Energy Use Assessments at Water and Wastewater Systems Guide
.
December 2012. Office of Water
,
Washington, DC
.
EPA (Environmental Protection Agency)
2013
Energy Efficiency in Water and Wastewater Facilities A Guide to Developing and Implementing Greenhouse Gas Reduction Programs, Washington, DC
.
Marchi
A.
Simpson
A. R.
2013
Correction of the EPANET inaccuracy in computing the efficiency of variable speed pumps
.
Journal of Water Resources Planning and Management
139
(
4
),
456
459
.
MOE (Market Observatory for Energy)
2011
Key Figures. Directorate General for Energy
.
European Commission
,
Brussels
.
NYSERDA
2008
Statewide Assessment of Energy Use by the Municipal Water and Wastewater Sector
.
New York State Energy Research and Development Authority
,
Albany, New York
.
Pardo
M. A.
Manzano
J.
Cabrera
E.
García-Serra
J.
2013
Energy audit of irrigation networks
.
Journal of Biosystems Engineering
105
,
89
101
.
Renaud
E.
Clauzier
M.
Sandraz
A. C.
Pillot
J.
Gilbert
D.
2014
Introducing pressure and number of connections into water loss indicators for French drinking water supply networks
.
Water Science & Technology: Water Supply
14
(
6
),
1105
1111
.
Rossman
L. A.
1993
EPANET Users Manual
.
EPA-600/R-94/057, Environmental Protection Agency, Risk Reduction Engineering Laboratory
,
Cincinnati
.
Rossman
L. A.
1999
The EPANET Programmer's Toolkit for Analysis of Water Distribution
.
Rossman
L. A.
2000
Epanet2 Users Manual
.
US EPA
,
Cincinnati
.
Rossman
L. A.
2011
An Overview of EPANET Version 3.0
.
Water Distribution Systems Analysis
2010
,
14
18
.
doi: 10.1061/41203(425)3
.
Water in the West (WW)
2013
Water and Energy Nexus: A Literature Review
.
Stanford University
,
Standford, CA
.
Wolff
G.
2010
Agua y energía en California (Water and energy in California)
.
Ingeniería del Agua
17
(
3
),
201
211
.