DEVELOPMENTS IN EHV/UHV CIRCUIT BREAKER TESTING

Transcription

1 DEVELOPMENTS IN EHV/HV CIRCIT BREAKER TESTING ABSTRACT R.P.P. Smeets, A.B. Hofstee, M. Dekker DNV GL, KEMA Laboratories trechtseweg 310, 6812 AR Arnhem, the Netherlands Nowadays, a clear trend of increase of system voltage to 800 kv AC and above can be observed in the world. This calls for the need of having test facilities to verify the current interruption performance of circuit breakers for such super grids. Recent work of CIGRE WG A3.22 and A3.28 led to standardized requirements for the highest system voltages up to 1200 kv, now anchored in the IEC requirements. The present contribution highlights the test-circuits and test-methods that DNV GL s KEMA Laboratories have developed in recent year to be able to carry out a full test-program of short-circuit tests on circuit breakers up to 1200 kv. For such tests, separate, well-synchronized sources for the supply of fault current and transient recovery voltage (TRV) are required, combined to synthetic test circuits. In this contribution, it is shown how this principle is stretched further to be applied for HV switchgear. A recently commissioned system of a two-stage synthetic installation will be demonstrated, with actual test examples of 800 and 1100 kv circuit breakers, for which TRVs in excess of 2000 kv are required. A key feature of the test-circuit is that the circuit breaker under test can remain on ground, which greatly facilitates efficient testing compared to test approaches in which the breaker must be installed on an isolated platform. Examples of short-circuit current making- and breaking test are given, including the full power testing of a switch, designed to activate closing resistors in an 800 kv GIS circuit breaker. The expansion of the laboratory enables full three-phase synthetic testing, even under effectively earthed conditions, examples of which will be shown. Also, an outlook will be given on the future testing capabilities. KEYWORDS ltra-high voltage, short-circuit, testing, circuit breaker 1. INTRODCTION Long distance transmission of electrical energy calls for the need of ever higher voltage transmission systems. The most recent example is the commissioning of a 1100 kv AC transmission system in China early 2009 and the planning of a 1200 kv AC system in India. The key protection component in such systems is the circuit-breaker, consisting of 2-4 interrupters in series. Its main function is to interrupt fault current, in case a fault occurs in the system, most likely on an overhead line. After the fault current has been interrupted at a power frequency current zero, the natural reaction of the power system is the generation of a transient recovery voltage (TRV) that arises across the circuit-breaker. It consists of a contribution from the upstream- and from downstream side of the breaker, since after current interruption and separation of the two circuit parts, up- and downstream side circuits generate transients completely independently. The challenge for the breaker is (1) initially the very fast rate of rise of TRV while the breaker is still recovering from the thermal arcing stresses before current zero and (2) later the dielectric stresses due to the high peak value of TRV, overshooting the power frequency crest voltage [1]. The process of fault current interruption and its associated transient recovery voltage is shown in the simplified sketch of fig. 1. In the case of fig. 1, a single frequency TRV is drawn. However, in practice a very complicated TRV wave-shape is generated depending on the network earthing, type and location of fault etc. After the decay of the transient, the breaker stress reduces to the power frequency Recovery Voltage (RV).

2 For the purpose of type testing, a standardized system of TRV description is developed by the International Electro-technical Commission (IEC). In the relevant IEC standard [2], TRV envelopes are described for a variety of fault switching duties. These TRV envelopes are simply two or three connected lines, each characterized by two parameters. Thus, a two- or a four-parameter TRV can be constructed easily [2]. The present IEC standard, thanks to the prestandardization efforts of CIGRE Working Groups A3.22 and A3.28 now covers the system voltages of 1100 and 1200 kv [3, 4]. Apart from producing the necessary fault current, test laboratories need to produce TRV in highpower test that are in full conformity with the IEC four-parameter envelopes. fault current power frequency voltage contact separation arcing tme current zero Recovery Voltage (RV) Transient Recovery Voltage (TRV) Fig.1: The current interruption process in a single phase situation This paper will mainly focus on the proper testing and realization of the TRV wave shapes for circuit-breakers of 800 kv and above, since this is the main challenge in testing these devices. Also, attention will be paid to the latest technology of three-phase synthetic testing, and conditions will be identified where three-phase testing cannot be replaced by single-phase testing. 2. SYNTHETIC TESTING OF HIGH-VOLTAGE CIRCIT BREAKERS Testing (ultra-)high voltage circuit-breakers is a challenge in itself, since no test facility has sufficient power available to provide current and voltage as during a fault in the high-voltage networks. Given a (three-phase) fault current of 63 ka in an 800 kv system, the short-circuit power is 87.3 GVA, this is approx. 6 times higher than the direct power that world's largest test laboratory (DNV GL KEMA laboratories) can provide. To overcome this, test-laboratories employ the synthetic test technology. The philosophy behind synthetic testing is the availability of sufficient current during the current period (especially during the arcing time) - in order to provide sufficient thermal stress to the breaker before interruption - and adequate TRV after interruption in order to provide realistic dielectrical stresses during the TRV and RV periods. Generally, the fault/arc current is provided by short-circuit generators (the current circuit), whereas TRV is supplied by a high-voltage circuit (the voltage circuit), energized by a pre-charged capacitor bank. For proper synchronization of TRV to be applied exactly at current zero, two methods are in use: 1) Current injection schemes or Weil-Dobke circuits [5], in which shortly before interruption the voltage circuit takes over the supply of arc current and supplies the complete TRV after interruption. This circuit is mostly used in synthetic testing because of its equivalence with the service situation and less-critical timing. 2) Voltage injection schemes, in which the voltage circuit only supplies the TRV. In both cases, at least one auxiliary breaker is required to insulate the current circuit from the voltage circuit at the moment TRV is applied. Synthetic testing of HV circuit-breakers is standardized in IEC [6]. Current distortion by the arc voltages All HV circuit-breakers are equipped with more than one interrupter (arcing chamber) in series to cope with the very high system voltage and transients. Full pole synthetic tests on HV circuit-breakers lead to test circuits with at least 4 interrupters (2-4 belonging to the test object and 2-4 to the auxiliary breaker) in series during the high-current interval. Since the current source supplies the arc current with a much lower voltage (typically several tens of kv) than in service (many hundreds of Fig. 2: Voltage stresses during TRV at current interruption of metal enclosed circuit-breakers. a) Half-pole test, b) full-pole test [10]

3 kv) the summed arc voltages in synthetic circuits could take up a considerable fraction of the supply voltage. This results in a reduction of short-circuit current and thus of arc energy during the arcing period, compared to the situation in service or during direct tests. The associated reduction of arc energy can ease the testing of puffer type CBs, but can reduce the extinction capability of arc assisted CBs (self-blast, self-compression etc.) [7]. In order to prevent current distortion, the supply voltage of the current source, and therefore the power of the current source, has to be large enough. Supply voltage in the range of 48 to 60 kv for HV test circuits are necessary for synthetic testing to be equivalent with the service situation when the fault current is energized by the ()HV power system. For testing circuit-breakers with very low arc voltage, such as vacuum breakers, this argument does not hold, and rather low-voltage sources, e.g. 1 kv, can be satisfactory. Arc duration A challenge in synthetic testing is the arc duration. Because the voltage source that drives the test-current is much lower than the system voltage, the breaker under test tends to interrupt the current already at an earlier current zero (after much shorter arc duration) than in service. Without additional measure, this would imply insufficient arc energy. Therefore, synthetic testing has to include a method to assure that realistic arc duration is achieved, even under lower driving voltage. Mostly, special arc prolongation circuits (in fig. 3 identified as ML) are applied that supply a steep current pulse just prior to current zero, hence forcing the arc to re-ignite and to continue during one or more power frequency loops. 3. TESTING OF METAL-ENCLOSED CIRCIT-BREAKERS Circuit-breakers for rated voltage > 550 kv always consist of multiple interruption chambers in series, because single chambers are not able to withstand the (recovery) voltage at that level. Also at voltages in the range kv, breakers often consist of multiple interruption chambers. Grading capacitors across each interruption chamber usually guarantee an "approximately equal" share of the voltage by each interruption chamber. For the E(xtra) and (ltra) High Voltage levels ( kv and > 800 kv respectively), instead of testing the complete pole of a circuit-breaker, separate interrupter units are often tested with the appropriate portion of the rated voltage ('unit-testing', or 'half-pole' testing if one pole consists of two interrupter units). Transient dielectric stresses at current interruption. Half-pole tests on metal-enclosed circuit-breakers do not represent the correct (full) dielectric stresses between live parts and enclosure, at least for the short-circuit current tests (see fig. 2a), since only half the system voltage (/2) is between live internal parts and the enclosure, whereas in service this voltage is the full system voltage (fig. 2b). The hot exhaust gases, produced by the circuit-breaker(s) during fault current interruption may also reduce the dielectric withstand capability of the space surrounding the arcing chambers (between poles, across the chamber, to the enclosure), see fig. 2. With GIS and dead-tank circuit-breakers, gas dynamic phenomena and the influence of (hot, ionized, contaminated) exhaust gas have to be taken into account with respect to the decision to perform unit- or full-pole tests and with respect to the decision to which side of the circuit-breaker the largest dielectric stress has to be applied. In order to meet the requirements for metal enclosed (GIS, dead tank) switchgear according to IEC [9], the correct dielectric stresses between live parts and enclosure must be guaranteed also during the interruption process. Consequently, for metal enclosed circuit-breakers any test other than on the complete circuit-breaker including the stresses on all interrupter units is considered technically incorrect. Grading capacitors In testing of circuit-breakers with grading capacitors, unit tests may not represent the transient stresses that occur due to unequal dielectrical stress of the arcing chambers. In unit tests, stresses on grading capacitors (such as occur in pre-strikes) and on the breaker chambers are not represented. This is not trivial, since CIGRE identified grading capacitors as a major contributor to circuit-breaker failures [8]. A safety margin of some percent is usually applied in unit testing. This is to include an unequal voltage distribution because of an unequal distribution of (stray-) capacitance of the breaker units. Because grading capacitors are normally much larger than these stray capacitances, the unequal voltage distribution is covered by a safety margin of a few percent of voltage above 50% (for a two-chamber breaker). In case of designs without grading, this small safety margin is no longer adequate. Half-pole tests, without taking into account the fore-mentioned stresses, give inadequate evidence for the correct performance of the test object in service. Full-pole test are closer to the actual network situation, (see fig. 2b) as recognized in the IEC Standard.

4 4. TWO-STAGE SYNTHETIC TESTING In order to avoid discussions regarding half-pole testing a new test-circuit is designed for full-pole short-circuit current interruption testing of circuit-breakers up to 1200 kv. In principle, a test-method can be used in which 50% of the voltage is applied at both terminals of the breaker, so that the full TRV is applied across the breaker. This implies that the breaker must be installed on an insulated platform. Especially for the very large breakers this is highly unwanted, so that from the beginning on, a solution has been sought in which the breaker under test can remain on ground (potential). In order to realize this, a two-stage synthetic solution is adopted in which a new synthetic installation provides the second stage of the necessary (4 parameter) TRV superimposed on the voltage wave shape from the existing synthetic installation. Fig. 3: Lay-out of test circuit for testing circuit-breakers for 800 kv (upper left) and 1200 kv (upper right). The electrical circuit is shown in fig. 3. When applying voltage injection for TRV representation, the very first (few tens of micro-seconds) initial TRV rise must be produced by the supply circuit (R 1, C 1, C d ). Then, after firing of the spark-gap (GP), the components Rh 1, Lh 1, Ch 0 and Ch 1 (first synthetic stage) provide the first part of the TRV, up to roughly 1000 kv, whereas a similar circuitry (second synthetic stage) adds the second part to the TRV in the second stage, triggered by a second spark gap. Reactor Lh 3 is tuned with Ch 0 to provide (part of) the power frequency recovery voltage. Circuitry ML provides forced re-ignition (by discharge of a pre-charged capacitor bank) at first (and sometimes) second arc current zero, in order to achieve TRV [kv] and current [x100 A] T100s TRV 1st stage TRV IEC envelope 0 2nd stage TRV current time [ms] Fig. 4: Oscillogram of double stage synthetic 4 parameter TRV for an 800 kv circuit-breakers including IEC envelope.

5 arcing times that are realistic for circuit-breakers operating in HV systems. Absence of such a circuitry would cause interruption already after a short arcing time, since the low supply side TRV allows interruption after a very short arcing time. This method makes 800 kv, 1100 kv and 1200 kv testing possible without overstressing the test building s bushings and components. As can be seen, the main feature of the circuit is its capability to perform full-pole synthetic tests with the test breaker (TO) remaining on ground potential. Another major advantage with respect to HV synthetic test schemes proposed earlier [11] is the need for only one auxiliary breaker (AB in fig. 3). Fig. 4 shows an example of a synthetic test at 100% of the rated short-circuit breaking current. sing the same principle, full-pole tests up to 1200 kv rated voltage were demonstrated. In fig. 5, an oscillogram is shown of test-duty T10. As can be seen, by proper timing of the two circuits relative to each other, also a two parameter (single frequency) TRV as required for the T10 duty can be constructed. In order to maintain an undistorted fault current during the arcing phase (8 arcs in series), a supply voltage of 60 kv is chosen. Calculations were performed to predict the overvoltages after re-ignition of the test-breaker. This is necessary, because in voltage injection, reignition current is low, resulting in early interruption by the test breaker. This, in turn, leaves the capacitor banks only partially discharged, with the risk of unrealistic voltages after re-ignition. Suitable protection measures against these escalating voltage effects are taken. Several components, like reactors, and capacitor banks, have to be raised above earth potential to withstand a significant voltage (> 2000 kv switching impulse type level) to earth, some were modified and a new set of triggered spark gaps was developed. Fig. 6 gives an impression of the new synthetic installation, showing some of the reactors for TRV tuning. Fig. 7 gives an impression of the testing of an 800 kv (dead-tank) circuitbreaker. TRV (kv), current (A) T10 TRV IEC reference lines current (10% of 50 ka) time (us) Fig. 5: Realized TRV of T10 test duty for circuit-breaker and IEC envelope for 1200 kv. Fig. 6: Partial view of the new synthetic installation Fig. 7: Test set-up of full-pole testing of an 800 kv deadtank circuit-breaker

6 Short-circuit making tests The standardized test-duty T100s(a) [2,6] prescribes a making Fig. 8: Test circuit for full-pole synthetic make tests for circuit breakers of 800 kv (TO) operation with and above. For symbols see text. a closing operation against the symmetrical rated short-circuit breaking current. The rationale of this test is in the verification of the circuit-breaker to withstand the pre-arcing phenomena from the moment of pre-strike until mechanical contact touch and latch. Especially in circuit-breakers at the highest voltage level, the duration of the pre-arc can be up to several milliseconds. During this pre-arcing period, the symmetrical current represents the highest energy input by the arc. Circuits are developed in order to perform such making test under full-pole conditions for 800 kv and above. Also for this test, synthetic testing is applied, with a typical circuit as outlined in Fig. 8. Fig. 9 shows the procedure of such a full-pole synthetic make test. Prior to such a test, the maximum pre-arcing time is verified by application of voltage across the closing gap, but without power frequency current after pre-strike. Then, Voltage [kv] and current [x100 A] voltage across contacts pre-strike current time [ms] Fig.9: Oscillogram of a full-power synthetic make test of an 800 kv 63 ka circuit-breaker. in the full-power test, voltage is applied across the gap at the maximum pre-arcing time. At the moment the gap breaks down and simultaneously a series of special triggered make-switches is activated that starts the power frequency current to flow through the circuit-breaker gap. The time delay of the make-switch must be below 300 µs. In order to reduce the high-current arcing wear of the make-switch, it is shunted by a circuit-breaker that closes as soon as possible. 5. THREE-PHASE SYNTHETIC TESTING Breakers are in a common enclosure In principle, the standards allow tests to be performed on a single-phase basis only in the case of circuit-breakers consisting of three independent single-pole devices. In any case, the type tests must verify the capability of a circuit-breaker to interrupt three-phase faults. Therefore, three-phase tests are preferred whenever possible. This is of particular importance for circuit-breakers having their three poles in a common enclosure and also for those with individually enclosed poles but operated by a common operating mechanism. The use of the three-phase test procedure is necessary to reproduce the requested stresses in terms of arcing window, asymmetry and duration of minor, major and extended loops on the three poles. Preferably, all the above stresses should be applied in the same test. If this is impossible, a multi-part test procedure may be the only available method. While single-phase synthetic test methods are well established, three-phase synthetic testing is rather new and technically challenging. The principle that the current is obtained from a current source and the recovery voltage after a chosen current zero from a voltage source, as it is in single-phase synthetic testing, is also applied in three-phase synthetic testing. These circuits use a three-phase current source and depending on the system earthing, two or three voltage sources. One voltage source provides the TRV for the first pole-to-clear. For non-effectively earthed systems (kpp = 1.5), another voltage source supplies the TRVs for the second and third poles-to-clear because these poles clear simultaneously. For effectively earthed systems (kpp = 1.3) the second and third poles-to-clear clear at different moments, so another two independent voltage sources are necessary.

7 A three-phase synthetic circuit with current injection in all three phases, each clearing separately is shown in fig. 10. It consists of the following components: a three-phase current source powered by generators G and power transformers PT to raise the voltage to avoid influence of the arc voltages on the current. MB are master breakers and MS the making switches; voltage sources VS1, VS2, VS3, each a full scale synthetic installation providing current injection to each pole; a three-pole auxiliary circuit-breaker (AB1) of similar rating as the tested breaker, in order to isolate the current circuit from the TRV voltages; a three-pole tested circuit-breaker (TO); and arc prolongation circuits (ML1-3) connected to each phase of the current circuit to prevent an early interruption by the tested circuitbreaker and to assure realistic arc duration. At present, in the authors laboratory, sufficient experience is gathered to operate such test circuits. Having available three independently controllable synthetic installations enables now three-phase testing of three-pole breakers (in common enclosure) for effectively earthed systems. An example of such a test result is shown in fig. 11. Breakers are operated by a common drive For single-phase testing, it is necessary to assure whether it is justified to subject only one circuitbreaker pole to a test instead of all three. The interrupting process is affected by G MB MS L PT R S T ML 1 LNRN ML 2 ML 3 R1C1 SA Cd AB1 TO Fig. 10: Three-phase synthetic circuit with current injection in all phases for kpp = 1.3. Fig. 11: Three-phase synthetic testing of a three-phase enclosed 145 kv metal enclosed high-voltage circuitbreaker with first pole-to- clear factor kpp = 1.3. Blue traces are short-circuit current, red traces show TRVs and green traces are injected currents from the three synthetic test installations. Cs: contact separation electrodynamic forces of the high current, by the gas exhaust from the adjacent poles, by the contact speed, by the transient pressure of the extinguishing medium and the combination of several of these factors occurring simultaneously. The energy output of the operating mechanism should be properly considered since the stresses resulting from single-phase operation require only a reduced operating force in case the operation mechanism is designed to drive all three poles. In other words, the design of the circuit-breaker and the interaction of poles are particularly important when the three poles are driven by a common mechanism. In this case, short-circuit performance can only be verified by three-phase testing. VS1 VS2 VS3 Rh1 Chd Ch1 Rh2 Ch2 Lh2 Rh2 Ch2 Lh2 Lh3 Lh3 Lh3 Lh1 Lh1 Lh1 GP GP GP Ch0 Ch0 Ch0 This is explained in fig. 12. In this figure, the course of the contact travel is given for a noload operation, single-phase, and three-phase breaking test on the same circuit-breaker. The increasing stress on the mechanism, comparing no-load operation (no arc in the interrupter) with single-phase breaking (arcing in only one pole) and three-phase breaking (arcs in three poles) is evident. It shows appreciable slowing down of Fig. 12: Measured contact travel curves in no-load, singlephase breaking and three-phase breaking tests

8 the contact travel in the three-phase arcing condition and associated pressure built-up in the SF6 interrupter. Thus, conditions in the single test are not severe enough. These interactions do not apply to circuit-breakers having independently operated poles. 6. TESTING OF CLOSING RESISTOR SWITCH Many circuit-breakers operating in the highest voltage classes are equipped with closing resistors, in order to reduce switching transients [1]. Such closing resistors are switched temporarily in the circuit, just before closing of the main circuit-breaker, see the schematic of fig. 13. This implies that the full pre-strike and its pre-arc upon closing have to be absorbed by the contact system of the closing resistor switch. Relevant test circuits for testing this function of the resistor closing switch were developed and successfully applied in a test circuit for metal enclosed 800 kv switchgear. Figure 13: Resistor switch (S) switching the closing resistor (R) Fig. 14. Test-circuit for testing 800 kv closing resistor switch, using three voltage sources (SY1, SY2a, SY2b) in series to create 925 kv rms across the resistor switch (TO). The main issue here is the very high AC voltage that has to be applied across the resistor switch, just before making, followed by the full resistor inrush current after pre-strike. This voltage is generated by superposition of three synthetic installations, of which fig. 14 gives a principle scheme. Fig. 15 shows a typical test result, where an AC Voltage (red) 500 kv/div / current (blue) 5 ka/div application of voltage across switch pre-strike of resistor switch 1375 kv Time [10 ms/div] current Fig. 15: Result of making test of 800 kv closing resistor switch against 975 kvrms voltage of 1375 kvpk was applied across the switch prior to pre-strike. After pre-strike, a current of 3000 A is supplied, first feeding the pre-arc in the switch which will last until galvanic touch of the switch contacts. In these tests it was not the intention to verify the thermal capability of the resistors. 7. SMMARY AND FTRE OTLOOK As stipulated in the standard, metal enclosed high-voltage breakers have to tested in a full-pole set-up, which means that unit or half-pole testing is technically incorrect. A new synthetic test-circuit for testing kv circuit-breakers regarding their fault interruption capability is presented. It basically consists of two cascaded voltage injection circuits that can produce TRV in excess of 2000 kv. By proper dimensioning of its electrical parameters and by adequate coupling and timing, all standardized TRVs can be realized in combination with all possible levels of short-circuit current. A major advantage of this circuit is that test-objects can remain at ground (potential). Full-pole testing with sufficient value of supply voltage must guarantee a high degree of equivalence with the situation in service. Having three separate synthetic installations available also metal enclosed circuit-breakers with three breakers in a common enclosure can now be type-tested for applications in effectively earthed neutral systems. Also circuit-breakers that rely on a single mechanism, operating all three poles, must be tested in a three-phase test set-up. Together with the latest 2016 laboratory expansion to a total of 6 short-circuit generators (15 GVA short-circuit power) and to a total of 10 short-circuit transformers (550 kv direct voltage), see fig. 16, testing of the key components of the future ()HV grids is possible [12], notably: Capacitive current switching tests in a direct circuits on full-pole basis up to 550 kv rated voltage (voltage factor 1.4) with current up to 500 A. Rated voltages kv can be handled half-pole in a direct circuit up to 500 A. For capacitive current above 500 A synthetic solutions are available; short-circuit current up to 120 ka in synthetic testing;

9 shunt reactor switching tests up to and including rated voltages of 800 kv in a full-pole configuration. Fig. 16: One of the two new generators (left) and four short-circuit transformers (right) during installation BIBLIOGRAPHY [1] R.P.P Smeets, A.L.J. Janssen et al., Switching in Electrical Transmission and Distribution Systems, John Wiley, ISBN , 2015 [2] International Standard IEC : High-Voltage Switchgear and Controlgear Part 100: High- Voltage Alternating-Current Circuit-Breakers, Edition 2.1, [3] CIGRE Working Group A3.22, Technical Requirements for Substation Equipment exceeding 800 kv, CIGRE Technical Brochure 362, [4] CIGRE Working Group A3.28, Switching Phenomena for EHV and HV Equipment, CIGRE Technical Brochure 570, [5] J. Biermans, The Weil circuit for testing of high-voltage circuit-breakers with very high interrupting capacities, CIGRE Report 102, [6] IEC High-Voltage Switchgear and Controlgear Part 101: Synthetic testing. Ed. 2.0, [7] L. van der Sluis, B.L. Sheng, "The Influence of the Arc Voltage in Synthetic Test Circuits", IEEE Trans. on Pow. Del., Vol. 10, No. 1, 995, pp , [8] CIGRE Technical Brochure 368: "Operating Environment of voltage grading Capacitors Applied to High Voltage Circuit-breakers", CIGRE WG A3.18, 2009 [9] IEC "High-voltage switchgear and controlgear Part 203: Gas-insulated metal-enclosed switchgear for rated voltages above 52 kv", Ed. 2.0, [10] M. Kapetanovic, "High voltage circuit-breakers", Sarajevo 2010, ISBN (issued by KEMA Laboratories), 2011 [11] B. Sheng, A New Synthetic Test Circuit for ltra-high-voltage Circuit-breakers", Ph.D. thesis, Delft niversity of Technology [12] R.P.P. Smeets, "Saveguarding the supergrid", IEEE Spectrum, Dec. 2015, pp