Energy Submetering at WRRFs October 2017 October 2017 Nancy - - PowerPoint PPT Presentation

we weftec 2017 | 90th Annual WEFTEC Conference

October 2017

Energy Submetering at WRRFs

October 2017

Nancy Andrews rews

Survey Data Indicates Operations Frequently Implement Energy Monitoring

2

16 114 5 3 10 27 103 5 4 12 39 70 10 6 18 48 66 13 3 13 67 27 16 9 25 DO Opti timi mizatio tion Energy Monito torin ing Digester Opti timizatio tion Other Secondary Treatmen ent t Changes Enhanced Primary Treatmen ent Not Implemente ted d Due to Other Barriers Not Implemente ted d Due to Lack of Financ ncial ial Viabi bili lity ty Not Applicabl ble or Techni nically ally Feasible ble Change Made Not Consider ered ed

Barriers to Operations Initiatives

3

0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00

Lack of support from decision maker Operational conservatism Lack of energy data New technologies lack track records Insufficient training and knowledge Insufficient operator response to high energy use Energy data are from various sources and have long lag times Lack of technical support Lack of operation staff time Energy efficiency is not considered sufficiently important We don't want to be the first to try a new operating strategy

Less than 5 mgd Greater than 5 mgd

Prioritizing permit compliance Going beyond compliance

Energy Culture Reduces Barriers to Operations Initiatives

4

0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 Lack of support from decision maker Operation conservatism Lack of energy data New technologies lack track of records Insufficient training and knowledge Slow response from operators to make adjustments when needed Energy data are from various sources and have long lag times Lack of technical support Lack of operation staff time Energy efficiency is not considered sufficiently important We don't want to be the first to try a new operating strategy Developed Energy Programs Undeveloped Energy Programs

Brown and Caldwell

5

Pump Energy

Brown and Caldwell | 90th Annual WEFTEC Conference

6

Pump Efficiency Degrades Due to Internal Recirculation

Brown and Caldwell | 90th Annual WEFTEC Conference

7

Pump suction Pump discharge Recirculation

Pump Wear Increases Power Consumption

Brown and Caldwell | 90th Annual WEFTEC Conference

8

Pump Efficiency Increased by Routine Maintenance

Brown and Caldwell | 90th Annual WEFTEC Conference

9

Example Pump Station Instrumentation Schematic

10

SCADA Data Tracks Energy Impairments

CSWEA | 5-12-10 11

Flow Power er Pressure essure Speed eed

Sympt mptom

• m

Likely y Root t Cause se High Power Pipe Obstructed or Pump Impairment High Pressure Pipe Obstructed Sagging Pump Curve Pump Wear or Impairment High Speed Line Obstructed or Pump Impairment High Speed without High Pressure Pump Impairment

y = 0.2717x2 + 0.4799x + 99.637 R² = 0.9775

150 170 190 210 230 250 270 290 310 330 350 12 14 16 18 20 22 24 26 28

Power er, kW Flow, mgd

4/21/10, 5/3/10 7/1/10 to 7/21/10 7/24/10 - 7/26/10 7/27/10 - 8/7/10 8/19/10 11/2/10 - 11/5/10

• Poly. (11/2/10 - 11/5/10)

Sample real-time flow rate

Pump power performance target for sample real-time flow

Benchmark Equation Derived from Power Data

12

Power Performance Compared to Power Target

13

y = 0.2717x2 + 0.4799x + 99.637 R² = 0.9775 150 170 190 210 230 250 270 290 310 330 350 12 14 16 18 20 22 24 26 28 Power, er, kW Flow,

• w, mgd

4/21/10, 5/3/10 7/1/10 to 7/21/10 7/24/10 - 7/26/10 7/27/10 - 8/7/10 8/19/10 11/2/10 - 11/5/10

• Poly. (11/2/10 - 11/5/10)

Sample real-time flow rate Percent of Target 235/215 = 109% Sample operating condition

Target Power Metric for Alarming Problems and Tracking Improvements Over Variable Flow Conditions

14

90% 95% 100% 105% 110% 115% 120%

4/22/2010 4/27/2010 5/2/2010 5/7/2010 5/12/2010 5/17/2010 5/22/2010 5/27/2010

Percen cent t of Target t - Pump mp Power

Pump 3 cleaned Target Pump Performance Goal Pump Power Alarm Setting Pump impairment returns

• Use flow vs. power curve to establish target conditions
• Compare current conditions to shop drawing data to

quantify wear conditions

• Optimize wear ring maintenance to restore efficiency
• Focus on largest pumps
• Bias run hours toward most efficient pumps
• Identify pump and forcemain obstructions
• Abnormally high discharge pressure (force main)
• Abnormally high speed or energy (pump)

Leveraging Pump Data for Energy Reduction

Brown and Caldwell | 90th Annual WEFTEC Conference

15

Aeration Blower Energy

Brown and Caldwell | 90th Annual WEFTEC Conference

16

Expected factors affecting blower power relative to ideal behavior

Brown and Caldwell | 90th Annual WEFTEC Conference

17

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% No Load 50% Load Design Load Blower er Power er Oxygen gen Demand nd Removal (CBO BOD5 and NH3)

“Ideal” relationship Blower turndown inefficiency Blower minimum airflow Mixing limits Minimum flow per diffuser

Actual data is typically very scattered and power demand is flat or “non-ideal”

Brown and Caldwell | 90th Annual WEFTEC Conference

18

MCES Metro plant appears to have a more pronounced slope

Brown and Caldwell | 90th Annual WEFTEC Conference

19

MCES Metro “slope” is actually driven by seasonal effects

Brown and Caldwell | 90th Annual WEFTEC Conference

20

Summer Nitrification Winter Carbonaceous

Non-ideal response across observed loads

Brown and Caldwell | 90th Annual WEFTEC Conference

21

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% No Load 50% Load Design Load Blower er Power er Oxygen gen Demand nd (CBO BOD5 and N)

Typical blower power modulation relative to load

Ideal blower power relationship to load Blower power less than expected at peak load

Non-ideal response divided between aeration control and blower efficiency effects

Brown and Caldwell | 90th Annual WEFTEC Conference

22

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% No Load 50% Load Design Load Blower er Power er Oxygen gen Demand nd (CBO BOD5 and N)

Non-ideal power due to inefficient blower part-load operation

Typical blower power modulation relative to load

Ideal blower power relationship to load

Quantifying blower modulation efficiency

Brown and Caldwell | 90th Annual WEFTEC Conference

23

𝑄𝑠𝑝𝑑𝑓𝑡𝑡 𝑁𝑝𝑒𝑣𝑚𝑏𝑢𝑗𝑝𝑜 𝐹𝑔𝑔𝑗𝑑𝑗𝑓𝑜𝑑𝑧 = 𝑇𝑚𝑝𝑞𝑓𝑏𝑑𝑢𝑣𝑏𝑚 𝑇𝑚𝑝𝑞𝑓𝑗𝑒𝑓𝑏𝑚 Actual blower Ideal blower

• Ideal slope assumes both variables rise proportionally
• Process Modulation Efficiency is 100% if process is behaving ideally

Statistical parameters show range of blower non-idealness and variability

Brown and Caldwell | 90th Annual WEFTEC Conference

24

MCES Metro Summer Littleton-Englewood Brightwater

𝐷𝑊 = 𝑇𝑢𝑏𝑜𝑒𝑏𝑠𝑒 𝐸𝑓𝑤𝑗𝑏𝑢𝑗𝑝𝑜 𝐵𝑤𝑓𝑠𝑏𝑕𝑓

Non-ideal response divided between to aeration control and blower efficiency effects

Brown and Caldwell | 90th Annual WEFTEC Conference

25

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% No Load 50% Load Design Load Blower er Power er Oxygen gen Demand nd (CBO BOD5 and N)

Typical blower power modulation relative to load

Non-ideal power due to aeration control limitations Ideal blower power relationship to load Non-ideal power due to inefficient blower part-load operation

Overall process efficiency: Nitrifying WRRFs

Brown and Caldwell | 90th Annual WEFTEC Conference

26

𝑄𝑠𝑝𝑑𝑓𝑡𝑡 𝑁𝑝𝑒𝑣𝑚𝑏𝑢𝑗𝑝𝑜 𝐹𝑔𝑔𝑗𝑑𝑗𝑓𝑜𝑑𝑧 = 𝑃2 𝐸𝑓𝑛𝑏𝑜𝑒 𝑇𝑚𝑝𝑞𝑓𝑏𝑑𝑢𝑣𝑏𝑚 𝑇𝑚𝑝𝑞𝑓𝑗𝑒𝑓𝑏𝑚

Boulder, CO No Ammonia Control Grand Rapids, MI South Side Ammonia Control

Overall process efficiency

Brown and Caldwell | 90th Annual WEFTEC Conference

27

Boulder Grand Rapids North and South NYC DEP Jamaica

Slightly higher response to peak load conditions at some plants

Brown and Caldwell | 90th Annual WEFTEC Conference

28

𝑀𝑝𝑏𝑒 𝑞𝑓𝑏𝑙𝑗𝑜𝑕 𝑔𝑏𝑑𝑢𝑝𝑠 = 99𝑢ℎ 𝑞𝑓𝑠𝑑𝑓𝑜𝑢𝑗𝑚𝑓 𝑚𝑝𝑏𝑒 𝐵𝑤𝑓𝑠𝑏𝑕𝑓 𝑚𝑝𝑏𝑒 𝐶𝑚𝑝𝑥𝑓𝑠 𝑞𝑓𝑏𝑙𝑗𝑜𝑕 𝑔𝑏𝑑𝑢𝑝𝑠 = 𝐶𝑚𝑝𝑥𝑓𝑠 𝑞𝑝𝑥𝑓𝑠 ∗ 𝐵𝑤𝑓𝑠𝑏𝑕𝑓 𝑐𝑚𝑝𝑥𝑓𝑠 𝑞𝑝𝑥𝑓𝑠

*on peak load days

Significant blower energy correlation parameters

Brown and Caldwell | 90th Annual WEFTEC Conference

29

Influent CBOD-normalized blower power can be used to benchmark energy performance

Brown and Caldwell | 90th Annual WEFTEC Conference

30

Impact of primary TSS removal efficiency

Brown and Caldwell | 90th Annual WEFTEC Conference

31

• Use flow vs. kW curve to determine blower process

modulation efficiency.

• Compare to “ideal” or proportional curve
• Poor modulation efficiency will limit savings from process
• ptimization
• Blower power isn’t proportional to flow or load, so resist

using normalizing parameters (kWh/MG or kWh/lb BOD)

• Benchmark against other blower systems using WE&RF

data

• Consider ammonia monitoring and control systems
• Monitor increases in blower power due to diffuser fouling

and optimize diffuser maintenance investiments

Brown and Caldwell | 90th Annual WEFTEC Conference

32

Leveraging Blower Data for Energy Reduction

Digestion and Dewatering Energy

Brown and Caldwell | 90th Annual WEFTEC Conference

33

Solids handling energy is significant, but difficult to benchmark due to variations in metering scope (equipment vs. building)

Brown and Caldwell | 90th Annual WEFTEC Conference

34

WRRF Aver erage e Flow w (MGD) D) Total tal Solid ids s Proc

• cess

essin ing Ener ergy (kWh/lb /lb CBOD5) Type e of Solids lids Stabi abilization lization A 34 0.20 Digestion B 15 0.28 Digestion C 54 0.11 Thermally conditioning Multiple hearth incineration E 60 0.14 Digestion G 42 0.12 None KB KB 18 0.21 Digestion KS KS 77 0.32 Digestion LE LE 22 0.27 Digestion MB MB 27 0.14 Digestion MN MN 176 0.21 Fluid bed incineration MS MS 24 0.51 Multiple hearth incineration

Minimal daily variations in dewatering and digestion energy

Brown and Caldwell | 90th Annual WEFTEC Conference

35

• Longer term trends important
• Optimize thickening to reduce centrifuge power
• Look at auxiliary systems like odor control and HVAC
• Secondary importance relative to more variable

processes

Solids Handling Energy Monitoring

Brown and Caldwell | 90th Annual WEFTEC Conference

36

Metering and Dashboard Examples

Brown and Caldwell | 90th Annual WEFTEC Conference

37

Utilize Existing Metering Devices

Brown and Caldwell

38

Variable Frequency Drive PLC / DCS

Twisted Shielded Pair Using Existing Analog Output on VFD or Modbus network connection

SCADA System Plant Network

Connect local blower power metering into SCADA Output power metering from existing VFDs

Add Inexpensive Monitoring Devices

Brown and Caldwell

39

Existing Plant Network Switch Cat5E Shark 200 SCADA System Plant Network

CTs and PTs OR…Amp Metering Only

Mock Equipment Energy Dashboard

40

Today 30-day Rolling Average

Large Aeration Blowers 102% 101% ID Fans 98% 95% Fluidizing Air Blowers 93% 95% Centrifuges 106% 106% Effluent Pumps 110% 101% RAS Pumps 93% 99% 408 Compressors 99% 99% Cake Pumps 100% 102% Flotation Recycle Pumps 101% 104% Selector Zone Mixers 104% 106%

Sample Plant-Wide SCADA DASHBOARD

Brown and Caldwell

41

Nashville Energy Dashboard Screen

Brown and Caldwell

42

Trend lines showing kW usage Total Plant kW Usage Site Plan highlighting areas/equipment that corresponds to circuit selected Combined circuits grouped per plant area Instantaneous kW usage per circuit with meter showing relative usage. Set points for each meter will be added

Thank You

Nancy Andrews nandrews@brwncald.com