scieee Science in your language
[en] (orig)

Simplified indirect estimation of pump flow discharge: An example from Serbia

Abstract

In the absence of a flowmeter or due to its inaccuracy, the flow rate at the discharge section of the pipeline following the observed pump can be roughly estimated if the pressure can be measured instead. To use the proposed procedure two main conditions should be achieved: (1) a manometer should be installed at the discharge pipeline between the pump and the flow regulation valve, and (2) the actual curve that relates pressure and flow for the observed pump unit should be known in advance. The described example is from Serbia, but it is of interest for any water pumping station with a submersible pump (installed in wells or tanks) where a limited number of adequate places for the measuring of flow are available (if any are available at all), but where the pressure at the discharge pipe of the observed pump can be measured. This simplified method can find applicability in installations in remote rural regions where limited resources are available. The results show that the calculated values of the flow obtained by the presented method deviate greatly in relation to the measured values provided by the portable ultrasonic flowmeter, up to 60% at one of the measuring points. However, in relation to the measured values provided by the permanently installed flowmeter the discrepancy is significantly lower (0.6-6.8%).

Read accessible full text

Simplified indirect estimation of pump flow discharge: An example from Serbia

Author: Rašić Amon, Milica
Publisher: MDPI
Year: 2021
DOI: 10.3390/w13060796
Source: https://dspace.vsb.cz/bitstreams/81fcd1f4-c4dd-4c4b-a115-8c45fc4c25b3/download
wa e
Technical No e
Simpli ied Indi ec Es ima ion o Pump Flow Discha ge:
An Example om Se bia
Milica Raši´c Amon 1,2 , Milan Radi´c 1, Zo an S aji´c 1,2, Dejan B ki´c 1,2,3,* and Pa el P aks 3


Ci a ion: Raši´c Amon, M.; Radi´c, M.;
S aji´c, Z.; B ki´c, D.; P aks, P.
Simpli ied Indi ec Es ima ion o
Pump Flow Discha ge: An Example
om Se bia. Wa e 2021,13, 796.
h ps://doi.o g/10.3390/w13060796
Academic Edi o s: S e ano Al isi and
Helena M. Ramos
Recei ed: 17 Decembe 2020
Accep ed: 10 Ma ch 2021
Published: 14 Ma ch 2021
Publishe ’s No e: MDPI s ays neu al
wi h ega d o ju isdic ional claims in
published maps and ins i u ional a il-
ia ions.
Copy igh : © 2021 by he au ho s.
Licensee MDPI, Basel, Swi ze land.
This a icle is an open access a icle
dis ibu ed unde he e ms and
condi ions o he C ea i e Commons
A ibu ion (CC BY) license (h ps://
c ea i ecommons.o g/licenses/by/
4.0/).
1Facul y o Elec onic Enginee ing, Uni e si y o Niš, 18000 Niš, Se bia; [email p o ec ed] (M.R.A.);
[email p o ec ed] (M.R.); [email p o ec ed] (Z.S.)
2Resea ch and De elopmen Cen e “IRC Al a ec”, 18000 Niš, Se bia
3
IT4Inno a ions, VSB—Technical Uni e si y o Os a a, 708 00 Os a a, Czech Republic; [email p o ec ed]
*Co espondence: [email p o ec ed] o [email p o ec ed]
Abs ac :
In he absence o a lowme e o due o i s inaccu acy, he low a e a he discha ge
sec ion o he pipeline ollowing he obse ed pump can be oughly es ima ed i he p essu e can be
measu ed ins ead. To use he p oposed p ocedu e wo main condi ions should be achie ed: (1) a
manome e should be ins alled a he discha ge pipeline be ween he pump and he low egula ion
al e, and (2) he ac ual cu e ha ela es p essu e and low o he obse ed pump uni should be
known in ad ance. The desc ibed example is om Se bia, bu i is o in e es o any wa e pumping
s a ion wi h a subme sible pump (ins alled in wells o anks) whe e a limi ed numbe o adequa e
places o he measu ing o low a e a ailable (i any a e a ailable a all), bu whe e he p essu e a he
discha ge pipe o he obse ed pump can be measu ed. This simpli ied me hod can ind applicabili y
in ins alla ions in emo e u al egions whe e limi ed esou ces a e a ailable. The esul s show ha
he calcula ed alues o he low ob ained by he p esen ed me hod de ia e g ea ly in ela ion o he
measu ed alues p o ided by he po able ul asonic lowme e , up o 60% a one o he measu ing
poin s. Howe e , in ela ion o he measu ed alues p o ided by he pe manen ly ins alled lowme e
he disc epancy is signi ican ly lowe (0.6–6.8%).
Keywo ds:
pumps; p essu e d op; low es ima ion; accu a e measu emen ; pipelines; wa e -
wo ks; lowme e
1. In oduc ion
In he absence o a lowme e o due o i s inaccu acy, he low a e a he discha ge
sec ion o he pipeline ollowing he obse ed pump can be oughly es ima ed i he wa e
p essu e can be measu ed ins ead. Ve y o en, i is p ac ically impossible o selec an
adequa e place o a lowme e due o he p esence o many elemen s in a wa e pumping
s a ion, such as al es and elbows whe e he pipeline changes i s di ec ion. The p oblem
o inco ec measu emen s o he discha ge low a e in wa e wo ks can be easily sol ed
i a sec ion o he pipe wi h an adequa e leng h wi h no low dis u bances is o eseen a
he discha ge side o he obse ed pump. To sol e he p oblem, in his p ac ical echnical
no e an al e na i e, simpli ied app oach is gi en o a ough es ima ion o he low a e
a he discha ge sec ion o he pipeline o hose pumps whe e he lowme e is absen o
when he eading om an exis ing lowme e is un eliable due o an un a o able pipeline
con igu a ion o due o any o he eason [
1
–
7
]. A pump is a machine ha is used o ans e
quan i ies o liquids om one place o ano he , and he e i s low-p essu e cha ac e is ic
is used o es ima e wa e low a e a he discha ge side o wa e pumping s a ions. The
p esen ed simpli ied app oach can be adap ed o any o he liquid, including c ude oil
whe e a special applica ion can be ound in o sho e ins alla ions wi h limi ed places
a ailable o measu emen .
In p ac ice, he e a e si ua ions when he e a e no lowme e s ins alled in pumping
s a ions (especially in emo e low-income egions), which is especially common when i
Wa e 2021,13, 796. h ps://doi.o g/10.3390/w13060796 h ps://www.mdpi.com/jou nal/wa e
Wa e 2021,13, 796 2 o 15
comes o pumping s a ions o smalle capaci y. In such cases, po able lowme e s a e used
o es ima e he low a e, i.e., he amoun o liquid pumped in a ce ain pe iod o ime, bu
choosing he place o hei ins alla ion is no easy.
In ce ain ins alla ions wi h pumps, o make measu emen s, i is easie and less cos ly
o ins all only a manome e (in al eady exis ing ins alla ions, his will p e en addi ional
cu ing o he pipeline o ins all a lowme e ). I he place a which a low me e should be
ins alled is no p ope ly chosen in he design phase p ope ly chosen in he design phase,
he p oblem can be sol ed by a adical app oach, which in ol es he exca a ion o a pa
o he pipeline ha is o he wise unde g ound. Howe e , his adical app oach in ol es
signi ican cos s and ca ies he isk o acciden s.
The p oposed app oach o low a e es ima ion has a gene al cha ac e and can be
applied in any pumping s a ion wi h a subme ged pump (in a ank o a well)
p o ided ha
:
1.
I has a pump uni wi h an in-ad ance known ac ual
H(Q)
cu e gained om lab-
o a o y es ing (possible disc epancy be ween he design sys em cu es a ailable
om manu ac u e s and he on-si e one is equen and he e o e such cu es om
ca alogues should be e-e alua ed i possible), and,
2. The p essu e can be measu ed a he discha geside o he obse ed pump.
Res ic ions include a lack o equipmen o egula e he speed o he o o s, and he
use o wo o mo e pumps in a pa allel con igu a ion, which means ha hei pe o mance
canno always be de e mined (especially ue when he non- e u n al es a he pumps
a e no ully igh ened). Such a si ua ion o en occu s in small emo e pumping s a ions
whe e he ope a ing egime is no ully espec ed ( he diame e o he suc ion pipe is no
always he same as he discha ge pipe diame e when, due o he isk o ca i a ion, a la ge
diame e o he suc ion pipe is usually used).
In he absence o pe manen measu emen s o low in wa e wo ks, o whe e accu acy
is doub ul, he mos commonly used p ac ice is o compa e he esul s ha can be ob ained
om se e al di e en po able con ol me e s ( hese a e, as a ule, po able ul asonic
me e s) [
8
–
16
]. We will also es ou app oach in his manne . As i will be p esen ed, he
esul s es ima ed ollowing he ecommenda ion om his p ac ical echnical no e a e mo e
eliable compa ed o hose ob ained using po able lowme e s in sys ems wi h a complex
geome y o pipeline ollowing he pump [17–21].
Based on one p ac ical example om Se bia, he heo e ical backg ound will be p e-
sen ed u he in he ex . This simpli ied me hod can ind applica ion in ins alla ions in
emo e, low-income egions whe e limi ed places o measu emen a e a ailable (i any a e
a ailable a all).
Simila ough app oxima e me hods o es ima ion o wa e low using local p essu e
loss measu emen s a e no known o he au ho s, bu inaccu acies o lowme e s a e no
a e [22,23].
2. The Desc ip ion o he P oposed App oach
To apply he simpli ied app oach p oposed in his wo k o low- a e es ima ion,
i is necessa y o ha e a pump wi h in-ad ance p ecisely de e mined
H(Q)
cu e and
o ha e a manome e ins alled a he discha ge side o he pump be o e he al e o
low egula ion [
24
,
25
]. I is no ad isable o ely on he ca alogued cha ac e is ics o he
pump, especially o olde pump models. E en wi h he pump uni s o he same ype
and manu ac u ed in he same yea , he ope a ing cha ac e is ics will no be he same
and will depend on hei eal condi ion such as he numbe o ope a ing hou s, he wea
o impelle s, he condi ion o bea ings, he numbe o o e hauls and losses in elec ic
mo o s, e c.
Figu e 1shows a gene al a angemen o elemen s equi ed o he applica ion o
he p oposed indi ec low a e es ima ion in a pump s a ion wi h subme sible pumps
( he lowme e s depic ed in ed a e no p esen a all o a e possibly inaccu a e due o
he unsui able place o placemen in a complex pipeline con igu a ion). The example
o he placemen o a po able lowme e shown in Figu e 1is a ypical example o an
Wa e 2021,13, 796 3 o 15
inapp op ia e posi ion in which a signi ican e o in low measu emen occu s (high
measu emen unce ain y).
Wa e 2021, 13, x FOR PEER REVIEW 3 o 16
lowme e s depic ed in ed a e no p esen a all o a e possibly inaccu a e due o he
unsui able place o placemen in a complex pipeline con igu a ion). The example o he
placemen o a po able lowme e shown in Figu e 1 is a ypical example o an inapp o-
p ia e posi ion in which a signi ican e o in low measu emen occu s (high measu e-
men unce ain y).
Figu e 1. The layou o he elemen s equi ed o low a e es ima ion.
Rega ding he scheme om Figu e 1, one should no o ge abou he p essu e loss
in he e ical sec ion o he pipeline— he p essu e measu ed a i s op is lowe han a
he pump head.
Taking in o accoun he complexi y o he pipeline, he equi alen cha ac e is ics o
he pump should be e alua ed h ough he analysis o he indi idual sys em o pipes on
which he pump is ins alled. In such a case, i is di icul o alk abou he compliance o
noncompliance o he measu ed cha ac e is ics o hose gi en by he pump manu ac u e .
In p ac ice, he e a e si ua ions whe e lowme e s a e no ins alled in pumping s a-
ions, which is especially common when i comes o pumping s a ions o smalle capaci y
(usually wi h wo pump uni s ha do no wo k in a pa allel con igu a ion, bu whe e one
o hem is used as a backup pump). In such cases, po able lowme e s a e used o es ima e
he low a e, i.e., he amoun o liquid pumped in a ce ain pe iod, while he choice o he
posi ion o hei ins alla ion is no easy. I he e a e no manholes p epa ed in ad ance, i
is necessa y o exca a e he g ound a ound he pipeline. In such si ua ions, his is no a
common p ac ice, bu a he he lowme e is placed in any accessible place.
Fo he low- a e es ima ion, he p essu e can be exp essed in [Pa] (in [ba ], whe e 1
ba = 105 Pa) o as a p essu e head in a uni o leng h [m], whe e hese wo o ms a e ela ed
as 𝑝 =𝜌·𝑔·𝐻, whe e 𝑝 is p essu e [Pa] measu ed by he manome e a he dis-
cha ge side (exac ly abo e he pump), 𝜌 is he densi y o he wa e ~1000 kg/m3, g is g a -
i a ional cons an ~9.81 m/sec2, while 𝐻 is p essu e head [m]. 𝐻 is he co ec ion in [m],
which depends on he heigh di e ence be ween he posi ion o he manome e and he
wa e le el in he ank. I he wa e le el in he ank is cons an du ing he expe imen , he
eloci y a he suc ion side is 𝑉 =0 [m/s]. The eloci y a he discha ge side 𝑉 de-
pends on he pipeline diame e and he low a e, which a e unknown a iables a he
beginning. This p oblem can be sol ed ei he by he implemen a ion o i e a i e analysis,
o by p ope choice o he posi ion o he manome e (as i is shown in Sec ion 3.2.1). Ne-
glec ing local losses om he poin o luid suc ion o he poin o p essu e measu emen ,
o al head H in [m] can be es ima ed using Equa ion (1):
Figu e 1. The layou o he elemen s equi ed o low a e es ima ion.
Rega ding he scheme om Figu e 1, one should no o ge abou he p essu e loss in
he e ical sec ion o he pipeline— he p essu e measu ed a i s op is lowe han a he
pump head.
Taking in o accoun he complexi y o he pipeline, he equi alen cha ac e is ics o
he pump should be e alua ed h ough he analysis o he indi idual sys em o pipes on
which he pump is ins alled. In such a case, i is di icul o alk abou he compliance o
noncompliance o he measu ed cha ac e is ics o hose gi en by he pump manu ac u e .
In p ac ice, he e a e si ua ions whe e lowme e s a e no ins alled in pumping s a ions,
which is especially common when i comes o pumping s a ions o smalle capaci y (usually
wi h wo pump uni s ha do no wo k in a pa allel con igu a ion, bu whe e one o hem
is used as a backup pump). In such cases, po able lowme e s a e used o es ima e he
low a e, i.e., he amoun o liquid pumped in a ce ain pe iod, while he choice o he
posi ion o hei ins alla ion is no easy. I he e a e no manholes p epa ed in ad ance, i
is necessa y o exca a e he g ound a ound he pipeline. In such si ua ions, his is no a
common p ac ice, bu a he he lowme e is placed in any accessible place.
Fo he low- a e es ima ion, he p essu e can be exp essed in [Pa] (in [ba ], whe e
1 ba = 10
5
Pa) o as a p essu e head in a uni o leng h [m], whe e hese wo o ms a e
ela ed as
pdis =ρ·g·H
, whe e
pdis
is p essu e [Pa] measu ed by he manome e a he
discha ge side (exac ly abo e he pump),
ρ
is he densi y o he wa e ~1000 kg/m
3
,gis
g a i a ional cons an ~9.81 m/sec
2
, while
H
is p essu e head [m].
Hg
is he co ec ion in
[m], which depends on he heigh di e ence be ween he posi ion o he manome e and
he wa e le el in he ank. I he wa e le el in he ank is cons an du ing he expe imen ,
he eloci y a he suc ion side is
Vsuc =
0 [m/s]. The eloci y a he discha ge side
Vdis
depends on he pipeline diame e and he low a e, which a e unknown a iables a he
beginning. This p oblem can be sol ed ei he by he implemen a ion o i e a i e analysis,
o by p ope choice o he posi ion o he manome e (as i is shown in Sec ion 3.2.1).
Neglec ing local losses om he poin o luid suc ion o he poin o p essu e measu emen ,
o al head H in [m] can be es ima ed using Equa ion (1):
H =pdis
ρ·g+Hg+V2
dis
2·g(1)
Wa e 2021,13, 796 4 o 15
To es he app oach o indi ec es ima ion o pump low discha ge p oposed he e, he
al e a he discha ge side o he pump should be ully open. Fu he , he al e needs o be
g adually closed in disc e e s eps, which changes he alues o he low and subsequen ly
o he p essu e. The e o e, o each achie ed s a iona y ope a ing poin , he alues o
he p essu e
pdis
om he manome e should be eco ded. Based on Equa ion (1) and
g aph in Figu e 2, i is possible o es ima e he low a e
Qes
a he discha ge side o he
obse ed pump. The cu e in Figu e 2 ela es he o al head
H
and he low a e
Q
and
has o be ob ained om he manu ac u e ’s ca alogue o p e e ably om a labo a o y
es in ad ance. The possibili y o disc epancy be ween he labo a o y cha ac e is ics and
cha ac e is ics om he ca alogue is some hing ha exis s in eali y, especially o an olde
pump, which has been in se ice o a long pe iod o ime and has been de ined by alid
echnical s anda ds (such as [
26
]). Depending on he ype o pump equi ed, mino o majo
de ia ions om he ac o y cha ac e is ic a e pe mi ed. The echnical s anda d, EN ISO
9906:2012 [
26
], allows ela i ely la ge disc epancies be ween he ope a ing cha ac e is ics
o he pumps he ca alogue cha ac e is ics.
Wa e 2021, 13, x FOR PEER REVIEW 4 o 16
𝐻=𝑝
𝜌·𝑔+𝐻+𝑉

2·𝑔 (1)
To es he app oach o indi ec es ima ion o pump low discha ge p oposed he e,
he al e a he discha ge side o he pump should be ully open. Fu he , he al e needs
o be g adually closed in disc e e s eps, which changes he alues o he low and subse-
quen ly o he p essu e. The e o e, o each achie ed s a iona y ope a ing poin , he al-
ues o he p essu e 𝑝 om he manome e should be eco ded. Based on Equa ion (1)
and g aph in Figu e 2, i is possible o es ima e he low a e 𝑄 a he discha ge side o
he obse ed pump. The cu e in Figu e 2 ela es he o al head 𝐻 and he low a e 𝑄
and has o be ob ained om he manu ac u e ’s ca alogue o p e e ably om a labo a o y
es in ad ance. The possibili y o disc epancy be ween he labo a o y cha ac e is ics and
cha ac e is ics om he ca alogue is some hing ha exis s in eali y, especially o an olde
pump, which has been in se ice o a long pe iod o ime and has been de ined by alid
echnical s anda ds (such as [26]). Depending on he ype o pump equi ed, mino o
majo de ia ions om he ac o y cha ac e is ic a e pe mi ed. The echnical s anda d, EN
ISO 9906:2012 [26], allows ela i ely la ge disc epancies be ween he ope a ing cha ac e -
is ics o he pumps he ca alogue cha ac e is ics.
Figu e 2. Indi ec de e mina ion o low 𝑄 a he measu ing poin based on he cu e, which
ela es o al head 𝐻 and low 𝑄.
The e o e, he co esponding low 𝑄 can be de e mined indi ec ly o each meas-
u ing poin om he known 𝐻(𝑄) cu e, (wi h sa is ac o y measu emen unce ain y) o
each measu ing poin o he p essu e 𝑝 ob ained om he manome e , as shown in
Figu e 2 ( ecalcula ed o 𝐻).
Figu e 2.
Indi ec de e mina ion o low
Qes
a he measu ing poin based on he cu e, which ela es o al head
H
and
low Q.
The e o e, he co esponding low
Qes
can be de e mined indi ec ly o each mea-
su ing poin om he known
H(Q)
cu e, (wi h sa is ac o y measu emen unce ain y)
o each measu ing poin o he p essu e
pdis
ob ained om he manome e , as shown in
Figu e 2( ecalcula ed o H).
The ollowing simple algo i hm, shown in Figu e 3, should be used o es ima e he
low a e using he p oposed app oach.
Wa e 2021,13, 796 5 o 15
Wa e 2021, 13, x FOR PEER REVIEW 5 o 16
The ollowing simple algo i hm, shown in Figu e 3, should be used o es ima e he
low a e using he p oposed app oach.
Figu e 3. A simple algo i hm o es ima ion o discha ge low.
3. A P ac ical Example
The pumping s a ion obse ed in which he me hod has been es ed has had a long
his o y o ques ionable measu emen s o low due o he complex geome y o he pipeline
wi h an insu icien leng h o pipe whe e a lowme e can be ins alled. Simpli ied diag am
o he measu ing pump s a ion is shown in Figu e 4 ( he op iew).
Figu e 4. Simpli ied diag am o he pump s a ion.
Due o a se ious mis ake du ing he cons uc ion p ocess, he main discha ge pipeline
made o s eel DN700 is no in acco dance wi h he o iginal p ojec and has h ee elbows
a sho dis ances. Pumps a e subme sed in he wa e ank, which is placed benea h he
loo o he pump s a ion. The wa e le el in he ank is held a a cons an alue. Du ing
he mode niza ion o he pump s a ion ha has been ca ied ou ecen ly, he old pumps
we e aken ou and eplaced wi h wo new subme sible pumps (G und os SP 270-3A G).
Figu e 3. A simple algo i hm o es ima ion o discha ge low.
3. A P ac ical Example
The pumping s a ion obse ed in which he me hod has been es ed has had a long
his o y o ques ionable measu emen s o low due o he complex geome y o he pipeline
wi h an insu icien leng h o pipe whe e a lowme e can be ins alled. Simpli ied diag am
o he measu ing pump s a ion is shown in Figu e 4( he op iew).
Wa e 2021, 13, x FOR PEER REVIEW 5 o 16
The ollowing simple algo i hm, shown in Figu e 3, should be used o es ima e he
low a e using he p oposed app oach.
Figu e 3. A simple algo i hm o es ima ion o discha ge low.
3. A P ac ical Example
The pumping s a ion obse ed in which he me hod has been es ed has had a long
his o y o ques ionable measu emen s o low due o he complex geome y o he pipeline
wi h an insu icien leng h o pipe whe e a lowme e can be ins alled. Simpli ied diag am
o he measu ing pump s a ion is shown in Figu e 4 ( he op iew).
Figu e 4. Simpli ied diag am o he pump s a ion.
Due o a se ious mis ake du ing he cons uc ion p ocess, he main discha ge pipeline
made o s eel DN700 is no in acco dance wi h he o iginal p ojec and has h ee elbows
a sho dis ances. Pumps a e subme sed in he wa e ank, which is placed benea h he
loo o he pump s a ion. The wa e le el in he ank is held a a cons an alue. Du ing
he mode niza ion o he pump s a ion ha has been ca ied ou ecen ly, he old pumps
we e aken ou and eplaced wi h wo new subme sible pumps (G und os SP 270-3A G).
Figu e 4. Simpli ied diag am o he pump s a ion.
Due o a se ious mis ake du ing he cons uc ion p ocess, he main discha ge pipeline
made o s eel DN700 is no in acco dance wi h he o iginal p ojec and has h ee elbows a
sho dis ances. Pumps a e subme sed in he wa e ank, which is placed benea h he loo
o he pump s a ion. The wa e le el in he ank is held a a cons an alue. Du ing he
mode niza ion o he pump s a ion ha has been ca ied ou ecen ly, he old pumps we e
aken ou and eplaced wi h wo new subme sible pumps (G und os SP 270-3A G). Be o e
he ins alla ion, one o he new pumps, ma ked as P1 in Figu e 4, had been addi ionally
es ed by he manu ac u e ’s labo a o y.

Wa e 2021,13, 796 6 o 15
To e i y he accu acy o he simpli ied app oach p oposed he e in p ac ice, low
measu emen s in he obse ed pumping s a ion using lowme e s we e pe o med (in
u he ex
Qmes
) and hen hose measu ed alues we e compa ed o he es ima ed alues
(in u he ex
Qes
). Besides, he exac cu e o he obse ed G und os SP 270-3A G, which
ela es o al head
H
and low
Q
was de e mined in he manu ac u e ’s labo a o y ( he
es ing o he pump in he manu ac u e ’s acili y is shown in Figu e 5, while he esul s a e
shown in Table 1). In he speci ic si ua ion p esen ed in his pape , he eal cha ac e is ics o
he pump eco ded in he labo a o y di e om hose om he manu ac u e ’s ca alogue.
Fu he mo e, some imes i is no possible o ha e a s able equency o 50 Hz, so he
ac ual cu e om he labo a o y, which ela es he o al head
H
and he low
Q
should be
ecalcula ed o he ac ual condi ions in he obse ed wa e pumping s a ion [
27
]. As we
will show in he ollowing case, he ope a ing equency was se o 49 Hz, o show how o
ecalcula e a cu e designed o 50 Hz.
Wa e 2021, 13, x FOR PEER REVIEW 6 o 16
Be o e he ins alla ion, one o he new pumps, ma ked as P1 in Figu e 4, had been addi-
ionally es ed by he manu ac u e ’s labo a o y.
To e i y he accu acy o he simpli ied app oach p oposed he e in p ac ice, low
measu emen s in he obse ed pumping s a ion using lowme e s we e pe o med (in u -
he ex 𝑄) and hen hose measu ed alues we e compa ed o he es ima ed alues
(in u he ex 𝑄). Besides, he exac cu e o he obse ed G und os SP 270-3A G,
which ela es o al head 𝐻 and low 𝑄 was de e mined in he manu ac u e ’s labo a o y
( he es ing o he pump in he manu ac u e ’s acili y is shown in Figu e 5, while he
esul s a e shown in Table 1). In he speci ic si ua ion p esen ed in his pape , he eal
cha ac e is ics o he pump eco ded in he labo a o y di e om hose om he manu-
ac u e ’s ca alogue. Fu he mo e, some imes i is no possible o ha e a s able equency
o 50 Hz, so he ac ual cu e om he labo a o y, which ela es he o al head 𝐻 and he
low 𝑄 should be ecalcula ed o he ac ual condi ions in he obse ed wa e pumping
s a ion [27]. As we will show in he ollowing case, he ope a ing equency was se o 49
Hz, o show how o ecalcula e a cu e designed o 50 Hz.
Figu e 5. Tes ing o SP 270-3A G pump in he labo a o y.
Table 1. Values o he cu e ha ela es o al head and low o he SP 270-3A G pump measu ed
in he labo a o y a 50 Hz and ecalcula ed o 49 Hz.
50 Hz
𝑄 [dm3/s] 0 11.67 59.88 75.84 90.08
𝐻 [m] 145.38 138.89 97.47 86.2 71.96
49 Hz
𝑄 [dm3/s] 0 11.44 58.68 74.32 88.28
𝐻 [m] 139.62 133.39 93.61 82.79 69.11
The ollowing equipmen was used o he es in he pumping s a ion:
1. A subme sible pumping uni G und os SP 270-3A G wi h a a ed powe o 110 kW,
Figu e 5. Tes ing o SP 270-3A G pump in he labo a o y.
Table 1.
Values o he cu e ha ela es o al head and low o he SP 270-3A G pump measu ed in
he labo a o y a 50 Hz and ecalcula ed o 49 Hz.
50 Hz
Q[dm3/s] 0 11.67 59.88 75.84 90.08
H[m] 145.38 138.89 97.47 86.2 71.96
49 Hz
Q[dm3/s] 0 11.44 58.68 74.32 88.28
H[m] 139.62 133.39 93.61 82.79 69.11
The ollowing equipmen was used o he es in he pumping s a ion:
1. A subme sible pumping uni G und os SP 270-3A G wi h a a ed powe o 110 kW,
2.
The low was measu ed using a pe manen ly ins alled ul asonic lowme e Siemens
SITRANS F a he main discha ge pipeline a he obse ed pump s a ion (addi ionally
a po able ul asonic lowme e TDSH100 was used),
Wa e 2021,13, 796 7 o 15
3.
The p essu e was measu ed using an IDROSCAN digi al manome e which was
ins alled on he discha ge side o he pump (addi ionally, an analogue manome e
was ins alled),
4.
The equency o he powe supply o he d i e mo o was egula ed using a Mi subishi
F740 egula o , and
5.
The low was egula ed using he al e ins alled a he s eel discha ge pipe DN 200
a e he pump P1.
All equipmen used in ou expe imen s has been p ope ly es ed using all common
echnical no ms and s anda ds equi ed o such de ices [28].
3.1. De e mina ion o Subs i u e Cha ac e is ics
The pu pose o his subsec ion is o show how he accu a e
H(Q)
cu e o he obse ed
G und os SP 270-3A G was de e mined.
3.1.1. Measu ing in he Manu ac u e ’s Labo a o y and he Recalcula ion o he Ac ual
F equency o Powe Supply
The measu ing in he manu ac u e ’s labo a o y:
The ac ual
H(Q)
cu e o he ob-
se ed pump SP 270-3A G was de e mined expe imen ally in he manu ac u e ’s labo a o y
and he esul s a e alid o powe supply equency o 50 Hz ( ela ed lows and he p es-
su e ob ained in disc e e s eps, which ha e been used o cons uc he cu e a e gi en in
Table 1).
Labo a o y es s a e expensi e, bu hey a e no manda o y and do no ep esen
wha he au ho s sugges in e e y si ua ion. In ce ain si ua ions, especially when i
comes o pumping uni s o highe powe , he p ice o such es s may be negligible when
compa ed o he possibili ies ha he use can ge wi h he help o such es s. The ac
is ha pump manu ac u e s usually do no ag ee o p o e he pe o mance o pumping
uni s a he ins alla ion si es o such pumps, because hey canno gua an ee he accu acy o
he measu ing equipmen in he pumping s a ions whe e he obse ed pumps ha e been
ins alled. Ins ead, hey p e e es ing in an independen o he manu ac u e ’s labo a o y.
Also, he cha ac e is ics o pump uni s change du ing se ice li e. In si ua ions whe e he e
a e lowme e s ins alled in pumping s a ions, i is no complica ed o pe iodically pe o m
measu emen s a he ins alla ion si e o check and/o co ec he ope a ing cha ac e is ics
o pumping uni s. This can usually be done as a pa o planned main enance ac i i ies in
he pumping s a ions so ha i does no inc ease unning cos s. This way, he deg ee o
de e io a ion o he ope a ing cha ac e is ics o he pumping uni s can be moni o ed. This
is pa icula ly impo an because hei e iciency can be educed o such an ex en ha i
would be necessa y o epai and/o o eplace hem.
The obse ed pumping s a ion in which he me hod was es ed has had a long his o y
o ques ionable measu ing o low due o he complex geome y o he pipeline wi h
insu icien leng h o pipe whe e a lowme e could be ins alled. The ins alled pumps a e
old and due o ha d exploi a ion, hei cha ac e is ics we e al e ed compa ed o hose om
he ca alogue. The e o e, he old pumps we e aken ou and eplaced wi h new pumps
which ha e been addi ionally es ed by he manu ac u e ’s labo a o y ( es ing o he pump
in he manu ac u e ’s acili y is shown in Figu e 5, while he esul s a e shown in Table 1).
Recalcula ion o he ac ual equency o he powe supply:
Fo an illus a ion o
he p oposed es ima ion o he low in gene al condi ions, du ing he es in he wa e
pumping s a ion whe e he pump has been ins alled, he equency has been se o 49 Hz
using a Mi subishi F740 egula o and consequen ly, he cu e, which ela es o al head
H
and low
Q
o such condi ions has been ecalcula ed. The co ec ed alues ha e been
calcula ed using Equa ion (2) and a e also p esen ed g aphically in Figu e 6(based on he
da a poin s om Table 1).
H49 =49
50 2·H50
Q49 =49
50 ·Q50 ), (2)
Wa e 2021,13, 796 8 o 15
Wa e 2021, 13, x FOR PEER REVIEW 8 o 16
calcula ed using Equa ion (2) and a e also p esen ed g aphically in Figu e 6 (based on he
da a poin s om Table 1).
𝐻 =49
50·𝐻
𝑄 =49
50·𝑄 ⎭
⎬
⎫
, (2)
S a is ical calcula ions in he shown case a e based on a limi ed numbe o measu e-
men poin s om Table 1. Al hough his may be insu icien o a ho ough analysis whe e
he numbe o hese poin s should be inc eased, o p ac ical analyses, i e poin s can be
su icien especially in cases whe e mo e da a is no a ailable such as in his case.
Figu e 6. The expe imen ally de e mined cu e which ela es p essu e head and low o he SP
270-3A G pump ecalcula ed o 49 Hz.
3.1.2. Reg ession Model and S a is ical Analysis
To acili a e he use o he da a om Table 1, he o al head and he low o he pump
SP 270-3A G in ou expe imen can be p esen ed g aphically as gi en in Figu e 6. The
co ec ed g aph o 49 Hz is also gi en in Figu e 6, while he ela ed cu e based on e-
g ession analysis is gi en in Equa ion (3).
To a oid eading om he cu e p esen ed in Figu e 6, i is be e o use he eg es-
sion o mula as gi en in Equa ion (3), which gi es he low as he unc ion o he o al
head o he obse ed pump; he da a om he labo a o y o 50 Hz and ecalcula ed al-
ues o 49 Hz a e in Table 1.
𝑄 =−0.0018·𝐻−0.8643·𝐻+156.77 (3)
The da a om he measu emen ha was used o p oducing o Equa ion (3) is gi en
in Table 1. The ela ion based on he inpu da a o Table 1 was i ed by eg ession analysis
using S a g aphics Cen u ion, a s a is ical s a is ics package ha pe o ms and explains,
in plain language, bo h basic and highly ad anced s a is ical unc ions. We ha e ied se -
e al eg ession models, and we es ed hem using Pea son co ela ion and R-Squa ed
alue coe icien o de e mina ion [29,30]:
Figu e 6.
The expe imen ally de e mined cu e which ela es p essu e head and low o he SP 270-3A G pump ecalcula ed
o 49 Hz.
S a is ical calcula ions in he shown case a e based on a limi ed numbe o measu e-
men poin s om Table 1. Al hough his may be insu icien o a ho ough analysis whe e
he numbe o hese poin s should be inc eased, o p ac ical analyses, i e poin s can be
su icien especially in cases whe e mo e da a is no a ailable such as in his case.
3.1.2. Reg ession Model and S a is ical Analysis
To acili a e he use o he da a om Table 1, he o al head and he low o he pump
SP 270-3A G in ou expe imen can be p esen ed g aphically as gi en in Figu e 6. The
co ec ed g aph o 49 Hz is also gi en in Figu e 6, while he ela ed cu e based on
eg ession analysis is gi en in Equa ion (3).
To a oid eading om he cu e p esen ed in Figu e 6, i is be e o use he eg ession
o mula as gi en in Equa ion (3), which gi es he low as he unc ion o he o al head o
he obse ed pump; he da a om he labo a o y o 50 Hz and ecalcula ed alues o 49
Hz a e in Table 1.
Qes −49 =−0.0018·H2−0.8643·H+156.77 (3)
The da a om he measu emen ha was used o p oducing o Equa ion (3) is gi en
in Table 1. The ela ion based on he inpu da a o Table 1was i ed by eg ession analysis
using S a g aphics Cen u ion, a s a is ical s a is ics package ha pe o ms and explains, in
plain language, bo h basic and highly ad anced s a is ical unc ions. We ha e ied se e al
eg ession models, and we es ed hem using Pea son co ela ion and R-Squa ed alue
coe icien o de e mina ion [29,30]:
•
The coe icien o de e mina ion R-Squa ed alue deno ed as R
2
equals o 1.0 would
indica e a pe ec i , while an R-Squa ed alue o 0.0 would indica e he opposi e case;
•
The Pea son co ela ion coe icien is a s a is ic ha measu es he linea co ela ion
be ween wo a iables and has a alue be ween +1 and
−
1, whe e a alue o +1 is
a o al posi i e linea co ela ion, 0 is no linea co ela ion, and
−
1 is o al nega i e
linea co ela ion.
Wa e 2021,13, 796 9 o 15
In ou pa icula case, a linea model is sligh ly mo e sui able o he poin s om
Table 1, bu we use he squa ed model because i is mo e gene al and sui able o mos
cases. Fo example, he linea model
Qes −49 =
180.147
−
1.26628
·H
, shows he In e cep
180.147 and he Slope
−
1.26628 as p edic o s. These a iables a e s a is ically signi ican
because hei p- alues equal 0.000, which a e < 0.05. I also means ha he e is a s a is ically
signi ican co ela ion be ween
Qes −49
and
H
a he 95% con idence le el (de ails o he
s a is ical analysis a e gi en in Table 2).
Table 2.
Compa ison o al e na i e models ob ained by he S a g aphics Cen u ion so wa e using
Pea son co ela ion and R-Squa ed alue coe icien o de e mina ion.
Model Co ela ion R-Squa ed
Linea −0.9991 99.82%
Squa ed-H −0.9980 99.60%
Squa ed-Qes −49 ecip ocal-H 0.9978 99.56%
Squa e oo -H −0.9977 99.53%
Loga i hmic-H −0.9945 98.91%
Squa ed-Qes −49 loga i hmic-H −0.9920 98.41%
Squa ed-Qes −49 squa e oo -H −0.9866 97.34%
Recip ocal-H 0.9822 96.47%
Squa ed-Qes −49 −0.9799 96.03%
Squa e oo -Qes −49 squa ed-H −0.9760 95.26%
Squa e oo -Qes −49 −0.9663 93.37%
Double squa ed −0.9643 92.98%
Double squa e oo −0.9592 92.01%
Squa e oo -Qes −49 loga i hmic-H −0.9504 90.33%
Squa e oo -Qes −49 ecip ocal-H −0.9269 85.92%
Besides, i is possible o use he poin s om Figu e 6and Table 1 alid o he equency
o powe supply o 50 Hz o cons uc he cu e and he ela ed unc ion, while in ha case
he alues o he measu ed low and he p essu e om he nex wo subsec ions whe e he
equency was 49 Hz, should be ecalcula ed o 50 Hz, see Equa ion (2).
3.2. Measu emen in he Obse ed Pumping S a ion
The measu emen s in he obse ed pumping s a ion we e pe o med o compa ison
wi h he es ima ed alues ob ained using he shown simpli ied me hod. The p essu e mea-
su emen in his subsec ion is used as one o he inpu pa ame e s and is essen ial o he
me hod o low es ima ion desc ibed in his echnical no e. Howe e , low measu emen s
in his subsec ion a e used only o compa ison. The equency o he powe supply du ing
he measu emen a he wa e pumping s a ion was se o 49 Hz.
3.2.1. P essu e Measu emen
The p essu e change, in he sec ion be ween he pump and he al e o he egu-
la ion o low, was moni o ed by IDROSCAN digi al manome e wi h measu ing ange
0–50 ba
(wi h a eading esolu ion o 0.01 ba and wi h a decla ed nonlinea i y
±
0.2%).
The IDROSCAN digi al manome e was placed di ec ly on he discha ge side o he pump,
oge he wi h he analogue manome e (Figu e 7). Because o he posi ion o he manome-
e s, some undesi able a ia ion o low can occu which can in oduce a ce ain e o o
p essu e measu emen due o incomple ely s abilized low. Howe e , bo h manome e s
showed he iden ical alues o he measu ed p essu e. I should be men ioned ha due o
he speci ic placemen o he manome e (i is moun ed in a special socke on he elbow o
he discha ge pipeline, exac ly abo e he pump), he wa e eloci y a he poin o measu -
ing is locally conside ed o be
Vdis ≈
0 [m/s]. Thus, he measu ed p essu e is somewha
highe han he eal s a ic p essu e and includes he dynamic p essu e. Due o he heigh
di e ence be ween he posi ion o he manome e and he wa e le el in he ank du ing
ou expe imen , co ec ion
Hg
= 1.6 m ( ela ed o Figu e 1) is in oduced in o Equa ion (1).